bibtex_url
stringlengths 41
50
| bibtext
stringlengths 693
2.88k
| abstract
stringlengths 0
2k
| authors
listlengths 1
45
| title
stringlengths 21
206
| id
stringlengths 7
16
| type
stringclasses 2
values | arxiv_id
stringlengths 9
12
⌀ |
|---|---|---|---|---|---|---|---|
https://aclanthology.org/2024.acl-short.30.bib | @inproceedings{ho-etal-2024-mtp,
title = "{MTP}: A Dataset for Multi-Modal Turning Points in Casual Conversations",
author = "Ho, Gia-Bao and
Tan, Chang and
Darban, Zahra and
Salehi, Mahsa and
Haf, Reza and
Buntine, Wray",
editor = "Ku, Lun-Wei and
Martins, Andre and
Srikumar, Vivek",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-short.30",
pages = "314--326",
abstract = "Detecting critical moments, such as emotional outbursts or changes in decisions during conversations, is crucial for understanding shifts in human behavior and their consequences. Our work introduces a novel problem setting focusing on these moments as turning points (TPs), accompanied by a meticulously curated, high-consensus, human-annotated multi-modal dataset. We provide precise timestamps, descriptions, and visual-textual evidence high-lighting changes in emotions, behaviors, perspectives, and decisions at these turning points. We also propose a framework, TPMaven, utilizing state-of-the-art vision-language models to construct a narrative from the videos and large language models to classify and detect turning points in our multi-modal dataset. Evaluation results show that TPMaven achieves an F1-score of 0.88 in classification and 0.61 in detection, with additional explanations aligning with human expectations.",
}
| Detecting critical moments, such as emotional outbursts or changes in decisions during conversations, is crucial for understanding shifts in human behavior and their consequences. Our work introduces a novel problem setting focusing on these moments as turning points (TPs), accompanied by a meticulously curated, high-consensus, human-annotated multi-modal dataset. We provide precise timestamps, descriptions, and visual-textual evidence high-lighting changes in emotions, behaviors, perspectives, and decisions at these turning points. We also propose a framework, TPMaven, utilizing state-of-the-art vision-language models to construct a narrative from the videos and large language models to classify and detect turning points in our multi-modal dataset. Evaluation results show that TPMaven achieves an F1-score of 0.88 in classification and 0.61 in detection, with additional explanations aligning with human expectations. | [
"Ho, Gia-Bao",
"Tan, Chang",
"Darban, Zahra",
"Salehi, Mahsa",
"Haf, Reza",
"Buntine, Wray"
]
| {MTP}: A Dataset for Multi-Modal Turning Points in Casual Conversations | acl-short.30 | Poster | 2111.09983v1 |
https://aclanthology.org/2024.acl-short.31.bib | @inproceedings{buder-grondahl-2024-parameter,
title = "What Does Parameter-free Probing Really Uncover?",
author = {Buder-Gr{\"o}ndahl, Tommi},
editor = "Ku, Lun-Wei and
Martins, Andre and
Srikumar, Vivek",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-short.31",
pages = "327--336",
abstract = "Supervised approaches to probing large language models (LLMs) have been criticized of using pre-defined theory-laden target labels. As an alternative, parameter-free probing constructs structural representations bottom-up via information derived from the LLM alone. This has been suggested to capture a genuine {``}LLM-internal grammar{''}. However, its relation to familiar linguistic formalisms remains unclear. I extend prior work on a parameter-free probing technique called perturbed masking applied to BERT, by comparing its results to the Universal Dependencies (UD) formalism for English. The results highlight several major discrepancies between BERT and UD, which lack correlates in linguistic theory. This raises the question of whether human grammar is the correct analogy to interpret BERT in the first place.",
}
| Supervised approaches to probing large language models (LLMs) have been criticized of using pre-defined theory-laden target labels. As an alternative, parameter-free probing constructs structural representations bottom-up via information derived from the LLM alone. This has been suggested to capture a genuine {``}LLM-internal grammar{''}. However, its relation to familiar linguistic formalisms remains unclear. I extend prior work on a parameter-free probing technique called perturbed masking applied to BERT, by comparing its results to the Universal Dependencies (UD) formalism for English. The results highlight several major discrepancies between BERT and UD, which lack correlates in linguistic theory. This raises the question of whether human grammar is the correct analogy to interpret BERT in the first place. | [
"Buder-Gr{\\\"o}ndahl, Tommi"
]
| What Does Parameter-free Probing Really Uncover? | acl-short.31 | Poster | 1301.0952v1 |
https://aclanthology.org/2024.acl-short.32.bib | @inproceedings{zhang-etal-2024-atlas,
title = "{ATLAS}: Improving Lay Summarisation with Attribute-based Control",
author = "Zhang, Zhihao and
Goldsack, Tomas and
Scarton, Carolina and
Lin, Chenghua",
editor = "Ku, Lun-Wei and
Martins, Andre and
Srikumar, Vivek",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-short.32",
pages = "337--345",
abstract = "Lay summarisation aims to produce summaries of scientific articles that are comprehensible to non-expert audiences. However, previous work assumes a one-size-fits-all approach, where the content and style of the produced summary are entirely dependent on the data used to train the model. In practice, audiences with different levels of expertise will have specific needs, impacting what content should appear in a lay summary and how it should be presented. Aiming to address this, we propose ATLAS, a novel abstractive summarisation approach that can control various properties that contribute to the overall {``}layness{''} of the generated summary using targeted control attributes. We evaluate ATLAS on a combination of biomedical lay summarisation datasets, where it outperforms state-of-the-art baselines using mainstream summarisation metrics.Additional analyses provided on the discriminatory power and emergent influence of our selected controllable attributes further attest to the effectiveness of our approach.",
}
| Lay summarisation aims to produce summaries of scientific articles that are comprehensible to non-expert audiences. However, previous work assumes a one-size-fits-all approach, where the content and style of the produced summary are entirely dependent on the data used to train the model. In practice, audiences with different levels of expertise will have specific needs, impacting what content should appear in a lay summary and how it should be presented. Aiming to address this, we propose ATLAS, a novel abstractive summarisation approach that can control various properties that contribute to the overall {``}layness{''} of the generated summary using targeted control attributes. We evaluate ATLAS on a combination of biomedical lay summarisation datasets, where it outperforms state-of-the-art baselines using mainstream summarisation metrics.Additional analyses provided on the discriminatory power and emergent influence of our selected controllable attributes further attest to the effectiveness of our approach. | [
"Zhang, Zhihao",
"Goldsack, Tomas",
"Scarton, Carolina",
"Lin, Chenghua"
]
| {ATLAS}: Improving Lay Summarisation with Attribute-based Control | acl-short.32 | Poster | 2406.05625v1 |
https://aclanthology.org/2024.acl-short.33.bib | @inproceedings{du-etal-2024-embspatial,
title = "{E}mb{S}patial-Bench: Benchmarking Spatial Understanding for Embodied Tasks with Large Vision-Language Models",
author = "Du, Mengfei and
Wu, Binhao and
Li, Zejun and
Huang, Xuanjing and
Wei, Zhongyu",
editor = "Ku, Lun-Wei and
Martins, Andre and
Srikumar, Vivek",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-short.33",
pages = "346--355",
abstract = "The recent rapid development of Large Vision-Language Models (LVLMs) has indicated their potential for embodied tasks. However, the critical skill of spatial understanding in embodied environments has not been thoroughly evaluated, leaving the gap between current LVLMs and qualified embodied intelligence unknown. Therefore, we construct EmbSpatial-Bench, a benchmark for evaluating embodied spatial understanding of LVLMs. The benchmark is automatically derived from embodied scenes and covers 6 spatial relationships from an egocentric perspective. Experiments expose the insufficient capacity of current LVLMs (even GPT-4V). We further present EmbSpatial-SFT, an instruction-tuning dataset designed to improve LVLMs{'} embodied spatial understanding.",
}
| The recent rapid development of Large Vision-Language Models (LVLMs) has indicated their potential for embodied tasks. However, the critical skill of spatial understanding in embodied environments has not been thoroughly evaluated, leaving the gap between current LVLMs and qualified embodied intelligence unknown. Therefore, we construct EmbSpatial-Bench, a benchmark for evaluating embodied spatial understanding of LVLMs. The benchmark is automatically derived from embodied scenes and covers 6 spatial relationships from an egocentric perspective. Experiments expose the insufficient capacity of current LVLMs (even GPT-4V). We further present EmbSpatial-SFT, an instruction-tuning dataset designed to improve LVLMs{'} embodied spatial understanding. | [
"Du, Mengfei",
"Wu, Binhao",
"Li, Zejun",
"Huang, Xuanjing",
"Wei, Zhongyu"
]
| {E}mb{S}patial-Bench: Benchmarking Spatial Understanding for Embodied Tasks with Large Vision-Language Models | acl-short.33 | Oral | 2406.05756v1 |
https://aclanthology.org/2024.acl-short.34.bib | @inproceedings{wretblad-etal-2024-understanding,
title = "Understanding the Effects of Noise in Text-to-{SQL}: An Examination of the {BIRD}-Bench Benchmark",
author = {Wretblad, Niklas and
Riseby, Fredrik and
Biswas, Rahul and
Ahmadi, Amin and
Holmstr{\"o}m, Oskar},
editor = "Ku, Lun-Wei and
Martins, Andre and
Srikumar, Vivek",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-short.34",
pages = "356--369",
abstract = "Text-to-SQL, which involves translating natural language into Structured Query Language (SQL), is crucial for enabling broad access to structured databases without expert knowledge. However, designing models for such tasks is challenging due to numerous factors, including the presence of noise, such as ambiguous questions and syntactical errors. This study provides an in-depth analysis of the distribution and types of noise in the widely used BIRD-Bench benchmark and the impact of noise on models. While BIRD-Bench was created to model dirty and noisy database values, it was not created to contain noise and errors in the questions and gold SQL queries. We found that noise in questions and gold queries are prevalent in the dataset, with varying amounts across domains, and with an uneven distribution between noise types. The presence of incorrect gold SQL queries, which then generate incorrect gold answers, has a significant impact on the benchmark{'}s reliability. Surprisingly, when evaluating models on corrected SQL queries, zero-shot baselines surpassed the performance of state-of-the-art prompting methods. We conclude that informative noise labels and reliable benchmarks are crucial to developing new Text-to-SQL methods that can handle varying types of noise.",
}
| Text-to-SQL, which involves translating natural language into Structured Query Language (SQL), is crucial for enabling broad access to structured databases without expert knowledge. However, designing models for such tasks is challenging due to numerous factors, including the presence of noise, such as ambiguous questions and syntactical errors. This study provides an in-depth analysis of the distribution and types of noise in the widely used BIRD-Bench benchmark and the impact of noise on models. While BIRD-Bench was created to model dirty and noisy database values, it was not created to contain noise and errors in the questions and gold SQL queries. We found that noise in questions and gold queries are prevalent in the dataset, with varying amounts across domains, and with an uneven distribution between noise types. The presence of incorrect gold SQL queries, which then generate incorrect gold answers, has a significant impact on the benchmark{'}s reliability. Surprisingly, when evaluating models on corrected SQL queries, zero-shot baselines surpassed the performance of state-of-the-art prompting methods. We conclude that informative noise labels and reliable benchmarks are crucial to developing new Text-to-SQL methods that can handle varying types of noise. | [
"Wretblad, Niklas",
"Riseby, Fredrik",
"Biswas, Rahul",
"Ahmadi, Amin",
"Holmstr{\\\"o}m, Oskar"
]
| Understanding the Effects of Noise in Text-to-{SQL}: An Examination of the {BIRD}-Bench Benchmark | acl-short.34 | Poster | 1911.09781v3 |
https://aclanthology.org/2024.acl-short.35.bib | @inproceedings{coelho-etal-2024-dwell,
title = "Dwell in the Beginning: How Language Models Embed Long Documents for Dense Retrieval",
author = "Coelho, Jo{\~a}o and
Martins, Bruno and
Magalhaes, Joao and
Callan, Jamie and
Xiong, Chenyan",
editor = "Ku, Lun-Wei and
Martins, Andre and
Srikumar, Vivek",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-short.35",
pages = "370--377",
abstract = "This study investigates the existence of positional biases in Transformer-based language models for text representation learning, particularly in the context of web document retrieval. We build on previous research that demonstrated loss of information in the middle of input sequences for causal language models, extending it to the domain of embedding learning. We examine positional biases at multiple stages of the training pipeline for an encoder-decoder neural retrieval model, namely language model pre-training, contrastive pre-training, and contrastive fine-tuning. Experiments with the MS-MARCO document collection reveal that after contrastive pre-training the model already generates embeddings that better capture the beginning of the input content, with fine-tuning further aggravating this effect.",
}
| This study investigates the existence of positional biases in Transformer-based language models for text representation learning, particularly in the context of web document retrieval. We build on previous research that demonstrated loss of information in the middle of input sequences for causal language models, extending it to the domain of embedding learning. We examine positional biases at multiple stages of the training pipeline for an encoder-decoder neural retrieval model, namely language model pre-training, contrastive pre-training, and contrastive fine-tuning. Experiments with the MS-MARCO document collection reveal that after contrastive pre-training the model already generates embeddings that better capture the beginning of the input content, with fine-tuning further aggravating this effect. | [
"Coelho, Jo{\\~a}o",
"Martins, Bruno",
"Magalhaes, Joao",
"Callan, Jamie",
"Xiong, Chenyan"
]
| Dwell in the Beginning: How Language Models Embed Long Documents for Dense Retrieval | acl-short.35 | Poster | 2404.04163v1 |
https://aclanthology.org/2024.acl-short.36.bib | @inproceedings{rabinovich-2024-thats,
title = "That{'}s Optional: A Contemporary Exploration of {``}that{''} Omission in {E}nglish Subordinate Clauses",
author = "Rabinovich, Ella",
editor = "Ku, Lun-Wei and
Martins, Andre and
Srikumar, Vivek",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-short.36",
pages = "378--385",
abstract = "The Uniform Information Density (UID) hypothesis posits that speakers optimize the communicative properties of their utterances by avoiding spikes in information, thereby maintaining a relatively uniform information profile over time. This paper investigates the impact of UID principles on syntactic reduction, specifically focusing on the optional omission of the connector {``}that{''} in English subordinate clauses. Building upon previous research, we extend our investigation to a larger corpus of written English, utilize contemporary large language models (LLMs) and extend the information-uniformity principles by the notion of entropy, to estimate the UID manifestations in the usecase of syntactic reduction choices.",
}
| The Uniform Information Density (UID) hypothesis posits that speakers optimize the communicative properties of their utterances by avoiding spikes in information, thereby maintaining a relatively uniform information profile over time. This paper investigates the impact of UID principles on syntactic reduction, specifically focusing on the optional omission of the connector {``}that{''} in English subordinate clauses. Building upon previous research, we extend our investigation to a larger corpus of written English, utilize contemporary large language models (LLMs) and extend the information-uniformity principles by the notion of entropy, to estimate the UID manifestations in the usecase of syntactic reduction choices. | [
"Rabinovich, Ella"
]
| That{'}s Optional: A Contemporary Exploration of {``}that{''} Omission in {E}nglish Subordinate Clauses | acl-short.36 | Poster | 2405.20833v1 |
https://aclanthology.org/2024.acl-short.37.bib | @inproceedings{an-etal-2024-large,
title = "Do Large Language Models Discriminate in Hiring Decisions on the Basis of Race, Ethnicity, and Gender?",
author = "An, Haozhe and
Acquaye, Christabel and
Wang, Colin and
Li, Zongxia and
Rudinger, Rachel",
editor = "Ku, Lun-Wei and
Martins, Andre and
Srikumar, Vivek",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-short.37",
pages = "386--397",
abstract = "We examine whether large language models (LLMs) exhibit race- and gender-based name discrimination in hiring decisions, similar to classic findings in the social sciences (Bertrand and Mullainathan, 2004). We design a series of templatic prompts to LLMs to write an email to a named job applicant informing them of a hiring decision. By manipulating the applicant{'}s first name, we measure the effect of perceived race, ethnicity, and gender on the probability that the LLM generates an acceptance or rejection email. We find that the hiring decisions of LLMs in many settings are more likely to favor White applicants over Hispanic applicants. In aggregate, the groups with the highest and lowest acceptance rates respectively are masculine White names and masculine Hispanic names. However, the comparative acceptance rates by group vary under different templatic settings, suggesting that LLMs{'} race- and gender-sensitivity may be idiosyncratic and prompt-sensitive.",
}
| We examine whether large language models (LLMs) exhibit race- and gender-based name discrimination in hiring decisions, similar to classic findings in the social sciences (Bertrand and Mullainathan, 2004). We design a series of templatic prompts to LLMs to write an email to a named job applicant informing them of a hiring decision. By manipulating the applicant{'}s first name, we measure the effect of perceived race, ethnicity, and gender on the probability that the LLM generates an acceptance or rejection email. We find that the hiring decisions of LLMs in many settings are more likely to favor White applicants over Hispanic applicants. In aggregate, the groups with the highest and lowest acceptance rates respectively are masculine White names and masculine Hispanic names. However, the comparative acceptance rates by group vary under different templatic settings, suggesting that LLMs{'} race- and gender-sensitivity may be idiosyncratic and prompt-sensitive. | [
"An, Haozhe",
"Acquaye, Christabel",
"Wang, Colin",
"Li, Zongxia",
"Rudinger, Rachel"
]
| Do Large Language Models Discriminate in Hiring Decisions on the Basis of Race, Ethnicity, and Gender? | acl-short.37 | Poster | 2406.10486v1 |
https://aclanthology.org/2024.acl-short.38.bib | @inproceedings{calabrese-etal-2024-explainability,
title = "Explainability and Hate Speech: Structured Explanations Make Social Media Moderators Faster",
author = {Calabrese, Agostina and
Neves, Leonardo and
Shah, Neil and
Bos, Maarten and
Ross, Bj{\"o}rn and
Lapata, Mirella and
Barbieri, Francesco},
editor = "Ku, Lun-Wei and
Martins, Andre and
Srikumar, Vivek",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-short.38",
pages = "398--408",
abstract = "Content moderators play a key role in keeping the conversation on social media healthy. While the high volume of content they need to judge represents a bottleneck to the moderation pipeline, no studies have explored how models could support them to make faster decisions. There is, by now, a vast body of research into detecting hate speech, sometimes explicitly motivated by a desire to help improve content moderation, but published research using real content moderators is scarce. In this work we investigate the effect of explanations on the speed of real-world moderators. Our experiments show that while generic explanations do not affect their speed and are often ignored, structured explanations lower moderators{'} decision making time by 7.4{\%}.",
}
| Content moderators play a key role in keeping the conversation on social media healthy. While the high volume of content they need to judge represents a bottleneck to the moderation pipeline, no studies have explored how models could support them to make faster decisions. There is, by now, a vast body of research into detecting hate speech, sometimes explicitly motivated by a desire to help improve content moderation, but published research using real content moderators is scarce. In this work we investigate the effect of explanations on the speed of real-world moderators. Our experiments show that while generic explanations do not affect their speed and are often ignored, structured explanations lower moderators{'} decision making time by 7.4{\%}. | [
"Calabrese, Agostina",
"Neves, Leonardo",
"Shah, Neil",
"Bos, Maarten",
"Ross, Bj{\\\"o}rn",
"Lapata, Mirella",
"Barbieri, Francesco"
]
| Explainability and Hate Speech: Structured Explanations Make Social Media Moderators Faster | acl-short.38 | Poster | 2406.04106v1 |
https://aclanthology.org/2024.acl-short.39.bib | @inproceedings{fang-etal-2024-born,
title = "Born Differently Makes a Difference: Counterfactual Study of Bias in Biography Generation from a Data-to-Text Perspective",
author = "Fang, Biaoyan and
Dinesh, Ritvik and
Dai, Xiang and
Karimi, Sarvnaz",
editor = "Ku, Lun-Wei and
Martins, Andre and
Srikumar, Vivek",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-short.39",
pages = "409--424",
abstract = "How do personal attributes affect biography generation? Addressing this question requires an identical pair of biographies where only the personal attributes of interest are different. However, it is rare in the real world. To address this, we propose a counterfactual methodology from a data-to-text perspective, manipulating the personal attributes of interest while keeping the co-occurring attributes unchanged. We first validate that the fine-tuned Flan-T5 model generates the biographies based on the given attributes. This work expands the analysis of gender-centered bias in text generation. Our results confirm the well-known bias in gender and also show the bias in regions, in both individual and its related co-occurring attributes in semantic machining and sentiment.",
}
| How do personal attributes affect biography generation? Addressing this question requires an identical pair of biographies where only the personal attributes of interest are different. However, it is rare in the real world. To address this, we propose a counterfactual methodology from a data-to-text perspective, manipulating the personal attributes of interest while keeping the co-occurring attributes unchanged. We first validate that the fine-tuned Flan-T5 model generates the biographies based on the given attributes. This work expands the analysis of gender-centered bias in text generation. Our results confirm the well-known bias in gender and also show the bias in regions, in both individual and its related co-occurring attributes in semantic machining and sentiment. | [
"Fang, Biaoyan",
"Dinesh, Ritvik",
"Dai, Xiang",
"Karimi, Sarvnaz"
]
| Born Differently Makes a Difference: Counterfactual Study of Bias in Biography Generation from a Data-to-Text Perspective | acl-short.39 | Poster | 2204.05879v1 |
https://aclanthology.org/2024.acl-short.40.bib | @inproceedings{yasser-etal-2024-sign,
title = "Sign Language Translation with Sentence Embedding Supervision",
author = "Yasser, Hamidullah and
Genabith, Josef and
Espa{\~n}a-Bonet, Cristina",
editor = "Ku, Lun-Wei and
Martins, Andre and
Srikumar, Vivek",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-short.40",
pages = "425--434",
abstract = "State-of-the-art sign language translation (SLT) systems facilitate the learning process through gloss annotations, either in an end2end manner or by involving an intermediate step. Unfortunately, gloss labelled sign language data is usually not available at scale and, when available, gloss annotations widely differ from dataset to dataset. We present a novel approach using sentence embeddings of the target sentences at training time that take the role of glosses. The new kind of supervision does not need any manual annotation but it is learned on raw textual data. As our approach easily facilitates multilinguality, we evaluate it on datasets covering German (PHOENIX-2014T) and American (How2Sign) sign languages and experiment with mono- and multilingual sentence embeddings and translation systems. Our approach significantly outperforms other gloss-free approaches, setting the new state-of-the-art for data sets where glosses are not available and when no additional SLT datasets are used for pretraining, diminishing the gap between gloss-free and gloss-dependent systems.",
}
| State-of-the-art sign language translation (SLT) systems facilitate the learning process through gloss annotations, either in an end2end manner or by involving an intermediate step. Unfortunately, gloss labelled sign language data is usually not available at scale and, when available, gloss annotations widely differ from dataset to dataset. We present a novel approach using sentence embeddings of the target sentences at training time that take the role of glosses. The new kind of supervision does not need any manual annotation but it is learned on raw textual data. As our approach easily facilitates multilinguality, we evaluate it on datasets covering German (PHOENIX-2014T) and American (How2Sign) sign languages and experiment with mono- and multilingual sentence embeddings and translation systems. Our approach significantly outperforms other gloss-free approaches, setting the new state-of-the-art for data sets where glosses are not available and when no additional SLT datasets are used for pretraining, diminishing the gap between gloss-free and gloss-dependent systems. | [
"Yasser, Hamidullah",
"Genabith, Josef",
"Espa{\\~n}a-Bonet, Cristina"
]
| Sign Language Translation with Sentence Embedding Supervision | acl-short.40 | Poster | 2210.06312v1 |
https://aclanthology.org/2024.acl-short.41.bib | @inproceedings{thielmann-etal-2024-stream,
title = "{STREAM}: Simplified Topic Retrieval, Exploration, and Analysis Module",
author = {Thielmann, Anton and
Reuter, Arik and
Weisser, Christoph and
Kant, Gillian and
Kumar, Manish and
S{\"a}fken, Benjamin},
editor = "Ku, Lun-Wei and
Martins, Andre and
Srikumar, Vivek",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-short.41",
pages = "435--444",
abstract = "Topic modeling is a widely used technique to analyze large document corpora. With the ever-growing emergence of scientific contributions in the field, non-technical users may often use the simplest available software module, independent of whether there are potentially better models available. We present a Simplified Topic Retrieval, Exploration, and Analysis Module (STREAM) for user-friendly topic modelling and especially subsequent interactive topic visualization and analysis. For better topic analysis, we implement multiple intruder-word based topic evaluation metrics. Additionally, we publicize multiple new datasets that can extend the so far very limited number of publicly available benchmark datasets in topic modeling. We integrate downstream interpretable analysis modules to enable users to easily analyse the created topics in downstream tasks together with additional tabular information.The code is available at the following link: https://github.com/AnFreTh/STREAM",
}
| Topic modeling is a widely used technique to analyze large document corpora. With the ever-growing emergence of scientific contributions in the field, non-technical users may often use the simplest available software module, independent of whether there are potentially better models available. We present a Simplified Topic Retrieval, Exploration, and Analysis Module (STREAM) for user-friendly topic modelling and especially subsequent interactive topic visualization and analysis. For better topic analysis, we implement multiple intruder-word based topic evaluation metrics. Additionally, we publicize multiple new datasets that can extend the so far very limited number of publicly available benchmark datasets in topic modeling. We integrate downstream interpretable analysis modules to enable users to easily analyse the created topics in downstream tasks together with additional tabular information.The code is available at the following link: https://github.com/AnFreTh/STREAM | [
"Thielmann, Anton",
"Reuter, Arik",
"Weisser, Christoph",
"Kant, Gillian",
"Kumar, Manish",
"S{\\\"a}fken, Benjamin"
]
| {STREAM}: Simplified Topic Retrieval, Exploration, and Analysis Module | acl-short.41 | Poster | 2203.04786v1 |
https://aclanthology.org/2024.acl-short.42.bib | @inproceedings{reddy-etal-2024-docfinqa,
title = "{D}oc{F}in{QA}: A Long-Context Financial Reasoning Dataset",
author = "Reddy, Varshini and
Koncel-Kedziorski, Rik and
Lai, Viet and
Krumdick, Michael and
Lovering, Charles and
Tanner, Chris",
editor = "Ku, Lun-Wei and
Martins, Andre and
Srikumar, Vivek",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-short.42",
pages = "445--458",
abstract = "For large language models (LLMs) to be effective in the financial domain {--} where each decision can have a significant impact {--} it is necessary to investigate realistic tasks and data. Financial professionals often interact with documents spanning hundreds of pages, but most financial research datasets only deal with short excerpts from these documents. To address this, we introduce a long-document financial QA task. We augment 7,437 questions from the existing FinQA dataset with full-document context, extending the average context length from under 700 words in FinQA to 123k words in DocFinQA. We conduct extensive experiments over retrieval-based QA pipelines and long-context language models. Based on our experiments, DocFinQA proves a significant challenge for even state-of-the-art systems. We also provide a case study on a subset of the longest documents in DocFinQA and find that models particularly struggle with these documents. Addressing these challenges may have a wide-reaching impact across applications where specificity and long-range contexts are critical, like gene sequences and legal document contract analysis. DocFinQA dataset is publicly accessible.",
}
| For large language models (LLMs) to be effective in the financial domain {--} where each decision can have a significant impact {--} it is necessary to investigate realistic tasks and data. Financial professionals often interact with documents spanning hundreds of pages, but most financial research datasets only deal with short excerpts from these documents. To address this, we introduce a long-document financial QA task. We augment 7,437 questions from the existing FinQA dataset with full-document context, extending the average context length from under 700 words in FinQA to 123k words in DocFinQA. We conduct extensive experiments over retrieval-based QA pipelines and long-context language models. Based on our experiments, DocFinQA proves a significant challenge for even state-of-the-art systems. We also provide a case study on a subset of the longest documents in DocFinQA and find that models particularly struggle with these documents. Addressing these challenges may have a wide-reaching impact across applications where specificity and long-range contexts are critical, like gene sequences and legal document contract analysis. DocFinQA dataset is publicly accessible. | [
"Reddy, Varshini",
"Koncel-Kedziorski, Rik",
"Lai, Viet",
"Krumdick, Michael",
"Lovering, Charles",
"Tanner, Chris"
]
| {D}oc{F}in{QA}: A Long-Context Financial Reasoning Dataset | acl-short.42 | Poster | 2401.10744v1 |
https://aclanthology.org/2024.acl-short.43.bib | @inproceedings{kargaran-etal-2024-masklid,
title = "{M}ask{LID}: Code-Switching Language Identification through Iterative Masking",
author = "Kargaran, Amir Hossein and
Yvon, Fran{\c{c}}ois and
Schuetze, Hinrich",
editor = "Ku, Lun-Wei and
Martins, Andre and
Srikumar, Vivek",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-short.43",
pages = "459--469",
abstract = "We present MaskLID, a simple, yet effective, code-switching (CS) language identification (LID) method. MaskLID does not require any training and is designed to complement current high-performance sentence-level LIDs. Sentence-level LIDs are classifiers trained on monolingual texts to provide single labels, typically using a softmax layer to turn scores into probabilities. However, in cases where a sentence is composed in both L1 and L2 languages, the LID classifier often only returns the dominant label L1. To address this limitation, MaskLID employs a strategy to mask text features associated with L1, allowing the LID to classify the text as L2 in the next round. This method uses the LID itself to identify the features that require masking and does not rely on any external resource. In this work, we explore the use of MaskLID for two open-source LIDs (GlotLID and OpenLID), that are both based on the FastText architecture. Code and demo are available at https://github.com/cisnlp/MaskLID.",
}
| We present MaskLID, a simple, yet effective, code-switching (CS) language identification (LID) method. MaskLID does not require any training and is designed to complement current high-performance sentence-level LIDs. Sentence-level LIDs are classifiers trained on monolingual texts to provide single labels, typically using a softmax layer to turn scores into probabilities. However, in cases where a sentence is composed in both L1 and L2 languages, the LID classifier often only returns the dominant label L1. To address this limitation, MaskLID employs a strategy to mask text features associated with L1, allowing the LID to classify the text as L2 in the next round. This method uses the LID itself to identify the features that require masking and does not rely on any external resource. In this work, we explore the use of MaskLID for two open-source LIDs (GlotLID and OpenLID), that are both based on the FastText architecture. Code and demo are available at https://github.com/cisnlp/MaskLID. | [
"Kargaran, Amir Hossein",
"Yvon, Fran{\\c{c}}ois",
"Schuetze, Hinrich"
]
| {M}ask{LID}: Code-Switching Language Identification through Iterative Masking | acl-short.43 | Poster | 2406.06263v1 |
https://aclanthology.org/2024.acl-short.44.bib | @inproceedings{liu-etal-2024-empirical,
title = "An Empirical Analysis on Large Language Models in Debate Evaluation",
author = "Liu, Xinyi and
Liu, Pinxin and
He, Hangfeng",
editor = "Ku, Lun-Wei and
Martins, Andre and
Srikumar, Vivek",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-short.44",
pages = "470--487",
abstract = "In this study, we investigate the capabilities and inherent biases of advanced large language models (LLMs) such as GPT-3.5 and GPT-4 in the context of debate evaluation. We discover that LLM{'}s performance exceeds humans and surpasses the performance of state-of-the-art methods fine-tuned on extensive datasets. We additionally explore and analyze biases present in LLMs, including positional bias, lexical bias, order bias, which may affect their evaluative judgments. Our findings reveal a consistent bias in both GPT-3.5 and GPT-4 towards the second candidate response presented, attributed to prompt design. We also uncover a lexical bias in both GPT-3.5 and GPT-4, especially when label sets carry connotations such as numerical or sequential, highlighting the critical need for careful label verbalizer selection in prompt design. Additionally, our analysis indicates a tendency of both models to favor the debate{'}s concluding side as the winner, suggesting an end-of-discussion bias.",
}
| In this study, we investigate the capabilities and inherent biases of advanced large language models (LLMs) such as GPT-3.5 and GPT-4 in the context of debate evaluation. We discover that LLM{'}s performance exceeds humans and surpasses the performance of state-of-the-art methods fine-tuned on extensive datasets. We additionally explore and analyze biases present in LLMs, including positional bias, lexical bias, order bias, which may affect their evaluative judgments. Our findings reveal a consistent bias in both GPT-3.5 and GPT-4 towards the second candidate response presented, attributed to prompt design. We also uncover a lexical bias in both GPT-3.5 and GPT-4, especially when label sets carry connotations such as numerical or sequential, highlighting the critical need for careful label verbalizer selection in prompt design. Additionally, our analysis indicates a tendency of both models to favor the debate{'}s concluding side as the winner, suggesting an end-of-discussion bias. | [
"Liu, Xinyi",
"Liu, Pinxin",
"He, Hangfeng"
]
| An Empirical Analysis on Large Language Models in Debate Evaluation | acl-short.44 | Poster | 2406.00050v2 |
https://aclanthology.org/2024.acl-short.45.bib | @inproceedings{zouhar-etal-2024-fine,
title = "Fine-Tuned Machine Translation Metrics Struggle in Unseen Domains",
author = "Zouhar, Vil{\'e}m and
Ding, Shuoyang and
Currey, Anna and
Badeka, Tatyana and
Wang, Jenyuan and
Thompson, Brian",
editor = "Ku, Lun-Wei and
Martins, Andre and
Srikumar, Vivek",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-short.45",
pages = "488--500",
abstract = "We introduce a new, extensive multidimensional quality metrics (MQM) annotated dataset covering 11 language pairs in the biomedical domain. We use this dataset to investigate whether machine translation (MT) metrics which are fine-tuned on human-generated MT quality judgements are robust to domain shifts between training and inference. We find that fine-tuned metrics exhibit a substantial performance drop in the unseen domain scenario relative to both metrics that rely on the surface form and pre-trained metrics that are not fine-tuned on MT quality judgments.",
}
| We introduce a new, extensive multidimensional quality metrics (MQM) annotated dataset covering 11 language pairs in the biomedical domain. We use this dataset to investigate whether machine translation (MT) metrics which are fine-tuned on human-generated MT quality judgements are robust to domain shifts between training and inference. We find that fine-tuned metrics exhibit a substantial performance drop in the unseen domain scenario relative to both metrics that rely on the surface form and pre-trained metrics that are not fine-tuned on MT quality judgments. | [
"Zouhar, Vil{\\'e}m",
"Ding, Shuoyang",
"Currey, Anna",
"Badeka, Tatyana",
"Wang, Jenyuan",
"Thompson, Brian"
]
| Fine-Tuned Machine Translation Metrics Struggle in Unseen Domains | acl-short.45 | Poster | 2402.18747v2 |
https://aclanthology.org/2024.acl-short.46.bib | @inproceedings{haq-etal-2024-indicirsuite,
title = "{I}ndic{IRS}uite: Multilingual Dataset and Neural Information Models for {I}ndian Languages",
author = "Haq, Saiful and
Sharma, Ashutosh and
Khattab, Omar and
Chhaya, Niyati and
Bhattacharyya, Pushpak",
editor = "Ku, Lun-Wei and
Martins, Andre and
Srikumar, Vivek",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-short.46",
pages = "501--509",
abstract = "In this paper, we introduce Neural Information Retrieval resources for 11 widely spoken Indian Languages (Assamese, Bengali, Gujarati, Hindi, Kannada, Malayalam, Marathi, Oriya, Punjabi, Tamil, and Telugu) from two major Indian language families (Indo-Aryan and Dravidian). These resources include (a) INDIC-MARCO, a multilingual version of the MS MARCO dataset in 11 Indian Languages created using Machine Translation, and (b) Indic-ColBERT, a collection of 11 distinct Monolingual Neural Information Retrieval models, each trained on one of the 11 languages in the INDIC-MARCO dataset. To the best of our knowledge, IndicIRSuite is the first attempt at building large-scale Neural Information Retrieval resources for a large number of Indian languages, and we hope that it will help accelerate research in Neural IR for Indian Languages. Experiments demonstrate that Indic-ColBERT achieves 47.47{\%} improvement in the MRR@10 score averaged over the INDIC-MARCO baselines for all 11 Indian languages except Oriya, 12.26{\%} improvement in the NDCG@10 score averaged over the MIRACL Bengali and Hindi Language baselines, and 20{\%} improvement in the MRR@100 Score over the Mr. Tydi Bengali Language baseline.",
}
| In this paper, we introduce Neural Information Retrieval resources for 11 widely spoken Indian Languages (Assamese, Bengali, Gujarati, Hindi, Kannada, Malayalam, Marathi, Oriya, Punjabi, Tamil, and Telugu) from two major Indian language families (Indo-Aryan and Dravidian). These resources include (a) INDIC-MARCO, a multilingual version of the MS MARCO dataset in 11 Indian Languages created using Machine Translation, and (b) Indic-ColBERT, a collection of 11 distinct Monolingual Neural Information Retrieval models, each trained on one of the 11 languages in the INDIC-MARCO dataset. To the best of our knowledge, IndicIRSuite is the first attempt at building large-scale Neural Information Retrieval resources for a large number of Indian languages, and we hope that it will help accelerate research in Neural IR for Indian Languages. Experiments demonstrate that Indic-ColBERT achieves 47.47{\%} improvement in the MRR@10 score averaged over the INDIC-MARCO baselines for all 11 Indian languages except Oriya, 12.26{\%} improvement in the NDCG@10 score averaged over the MIRACL Bengali and Hindi Language baselines, and 20{\%} improvement in the MRR@100 Score over the Mr. Tydi Bengali Language baseline. | [
"Haq, Saiful",
"Sharma, Ashutosh",
"Khattab, Omar",
"Chhaya, Niyati",
"Bhattacharyya, Pushpak"
]
| {I}ndic{IRS}uite: Multilingual Dataset and Neural Information Models for {I}ndian Languages | acl-short.46 | Poster | 2312.09508v1 |
https://aclanthology.org/2024.acl-short.47.bib | @inproceedings{zhao-etal-2024-agr,
title = "{AGR}: Reinforced Causal Agent-Guided Self-explaining Rationalization",
author = "Zhao, Yunxiao and
Wang, Zhiqiang and
Li, Xiaoli and
Liang, Jiye and
Li, Ru",
editor = "Ku, Lun-Wei and
Martins, Andre and
Srikumar, Vivek",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-short.47",
pages = "510--518",
abstract = "Most existing rationalization approaches are susceptible to degeneration accumulation due to a lack of effective control over the learning direction of the model during training. To address this issue, we propose a novel approach AGR (\textbf{A}gent-\textbf{G}uided \textbf{R}ationalization), guiding the next action of the model based on its current training state. Specifically, we introduce causal intervention calculus to quantify the causal effects inherent during rationale training, and utilize reinforcement learning process to refine the learning bias of them. Furthermore, we pretrain an agent within this reinforced causal environment to guide the next step of the model. We \textit{theoretically} demonstrate that a good model needs the desired guidance, and \textit{empirically} show the effectiveness of our approach, outperforming existing state-of-the-art methods on BeerAdvocate and HotelReview datasets.",
}
| Most existing rationalization approaches are susceptible to degeneration accumulation due to a lack of effective control over the learning direction of the model during training. To address this issue, we propose a novel approach AGR (\textbf{A}gent-\textbf{G}uided \textbf{R}ationalization), guiding the next action of the model based on its current training state. Specifically, we introduce causal intervention calculus to quantify the causal effects inherent during rationale training, and utilize reinforcement learning process to refine the learning bias of them. Furthermore, we pretrain an agent within this reinforced causal environment to guide the next step of the model. We \textit{theoretically} demonstrate that a good model needs the desired guidance, and \textit{empirically} show the effectiveness of our approach, outperforming existing state-of-the-art methods on BeerAdvocate and HotelReview datasets. | [
"Zhao, Yunxiao",
"Wang, Zhiqiang",
"Li, Xiaoli",
"Liang, Jiye",
"Li, Ru"
]
| {AGR}: Reinforced Causal Agent-Guided Self-explaining Rationalization | acl-short.47 | Poster | 2306.14115v2 |
https://aclanthology.org/2024.acl-short.48.bib | @inproceedings{ranathunga-etal-2024-shoulders,
title = "Shoulders of Giants: A Look at the Degree and Utility of Openness in {NLP} Research",
author = "Ranathunga, Surangika and
De Silva, Nisansa and
Jayakody, Dilith and
Fernando, Aloka",
editor = "Ku, Lun-Wei and
Martins, Andre and
Srikumar, Vivek",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-short.48",
pages = "519--529",
abstract = "We analysed a sample of NLP research papers archived in ACL Anthology as an attempt to quantify the degree of openness and the benefit of such an open culture in the NLP community. We observe that papers published in different NLP venues show different patterns related to artefact reuse. We also note that more than 30{\%} of the papers we analysed do not release their artefacts publicly. Further, we observe a wide language-wise disparity in publicly available NLP-related artefacts.",
}
| We analysed a sample of NLP research papers archived in ACL Anthology as an attempt to quantify the degree of openness and the benefit of such an open culture in the NLP community. We observe that papers published in different NLP venues show different patterns related to artefact reuse. We also note that more than 30{\%} of the papers we analysed do not release their artefacts publicly. Further, we observe a wide language-wise disparity in publicly available NLP-related artefacts. | [
"Ranathunga, Surangika",
"De Silva, Nisansa",
"Jayakody, Dilith",
"Fern",
"o, Aloka"
]
| Shoulders of Giants: A Look at the Degree and Utility of Openness in {NLP} Research | acl-short.48 | Poster | 2406.06021v1 |
https://aclanthology.org/2024.acl-short.49.bib | @inproceedings{siegel-etal-2024-probabilities,
title = "The Probabilities Also Matter: A More Faithful Metric for Faithfulness of Free-Text Explanations in Large Language Models",
author = "Siegel, Noah and
Camburu, Oana-Maria and
Heess, Nicolas and
Perez-Ortiz, Maria",
editor = "Ku, Lun-Wei and
Martins, Andre and
Srikumar, Vivek",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-short.49",
pages = "530--546",
abstract = "In order to oversee advanced AI systems, it is important to understand their reasons for generating a given output. When prompted, large language models (LLMs) can provide natural language explanations or reasoning traces that sound plausible and receive high ratings from human annotators. However, it is unclear to what extent these explanations are truly capturing the factors responsible for the model{'}s predictions: the most {``}human-like{''} explanation may be different from the one that is most faithful to the model{'}s true decision making process. In this work, we introduce the correlational counterfactual test (CCT), a faithfulness metric based on counterfactual input edits that takes into account not just the binary label change, but the total shift in the model{'}s predicted label distribution. We evaluate the faithfulness of free-text explanations generated by few-shot-prompted LLMs from the Llama-2 family on three NLP tasks. We find that these explanations are indeed more likely to mention factors when they are impactful to the model{'}s prediction, with the degree of association increasing with model size but varying significantly by task.",
}
| In order to oversee advanced AI systems, it is important to understand their reasons for generating a given output. When prompted, large language models (LLMs) can provide natural language explanations or reasoning traces that sound plausible and receive high ratings from human annotators. However, it is unclear to what extent these explanations are truly capturing the factors responsible for the model{'}s predictions: the most {``}human-like{''} explanation may be different from the one that is most faithful to the model{'}s true decision making process. In this work, we introduce the correlational counterfactual test (CCT), a faithfulness metric based on counterfactual input edits that takes into account not just the binary label change, but the total shift in the model{'}s predicted label distribution. We evaluate the faithfulness of free-text explanations generated by few-shot-prompted LLMs from the Llama-2 family on three NLP tasks. We find that these explanations are indeed more likely to mention factors when they are impactful to the model{'}s prediction, with the degree of association increasing with model size but varying significantly by task. | [
"Siegel, Noah",
"Camburu, Oana-Maria",
"Heess, Nicolas",
"Perez-Ortiz, Maria"
]
| The Probabilities Also Matter: A More Faithful Metric for Faithfulness of Free-Text Explanations in Large Language Models | acl-short.49 | Oral | 2404.03189v2 |
https://aclanthology.org/2024.acl-short.50.bib | @inproceedings{testoni-etal-2024-naming,
title = "Naming, Describing, and Quantifying Visual Objects in Humans and {LLM}s",
author = "Testoni, Alberto and
Sprott, Juell and
Pezzelle, Sandro",
editor = "Ku, Lun-Wei and
Martins, Andre and
Srikumar, Vivek",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-short.50",
pages = "547--557",
abstract = "While human speakers use a variety of different expressions when describing the same object in an image, giving rise to a distribution of plausible labels driven by pragmatic constraints, the extent to which current Vision {\&} Language Large Language Models (VLLMs) can mimic this crucial feature of language use is an open question. This applies to common, everyday objects, but it is particularly interesting for uncommon or novel objects for which a category label may be lacking or fuzzy. Furthermore, similar patterns of variation are observed among human speakers for highly context-sensitive expressions, such as the quantifiers {`}few{'} or {`}most{'}. In our work, we evaluate VLLMs (FROMAGe, BLIP-2, LLaVA) on three categories (nouns, attributes, and quantifiers) where humans show great subjective variability concerning the distribution over plausible labels, using datasets and resources mostly under-explored in previous work. Our results reveal mixed evidence on the ability of VLLMs to capture human naming preferences at generation time: while some models are good at mimicking human distributions for nouns and attributes, all of them fail to assign quantifiers, a task that requires more accurate, high-level reasoning.",
}
| While human speakers use a variety of different expressions when describing the same object in an image, giving rise to a distribution of plausible labels driven by pragmatic constraints, the extent to which current Vision {\&} Language Large Language Models (VLLMs) can mimic this crucial feature of language use is an open question. This applies to common, everyday objects, but it is particularly interesting for uncommon or novel objects for which a category label may be lacking or fuzzy. Furthermore, similar patterns of variation are observed among human speakers for highly context-sensitive expressions, such as the quantifiers {`}few{'} or {`}most{'}. In our work, we evaluate VLLMs (FROMAGe, BLIP-2, LLaVA) on three categories (nouns, attributes, and quantifiers) where humans show great subjective variability concerning the distribution over plausible labels, using datasets and resources mostly under-explored in previous work. Our results reveal mixed evidence on the ability of VLLMs to capture human naming preferences at generation time: while some models are good at mimicking human distributions for nouns and attributes, all of them fail to assign quantifiers, a task that requires more accurate, high-level reasoning. | [
"Testoni, Alberto",
"Sprott, Juell",
"Pezzelle, S",
"ro"
]
| Naming, Describing, and Quantifying Visual Objects in Humans and {LLM}s | acl-short.50 | Poster | 2403.06935v3 |
https://aclanthology.org/2024.acl-short.51.bib | @inproceedings{leidinger-etal-2024-llms,
title = "Are {LLM}s classical or nonmonotonic reasoners? Lessons from generics",
author = "Leidinger, Alina and
Van Rooij, Robert and
Shutova, Ekaterina",
editor = "Ku, Lun-Wei and
Martins, Andre and
Srikumar, Vivek",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-short.51",
pages = "558--573",
abstract = "Recent scholarship on reasoning in LLMs has supplied evidence of impressive performance and flexible adaptation to machine generated or human critique. Nonmonotonic reasoning, crucial to human cognition for navigating the real world, remains a challenging, yet understudied task. In this work, we study nonmonotonic reasoning capabilities of seven state-of-the-art LLMs in one abstract and one commonsense reasoning task featuring generics, such as {`}Birds fly{'}, and exceptions, {`}Penguins don{'}t fly{'} (see Fig. 1). While LLMs exhibit reasoning patterns in accordance with human nonmonotonic reasoning abilities, they fail to maintain stable beliefs on truth conditions of generics at the addition of supporting examples ({`}Owls fly{'}) or unrelated information ({`}Lions have manes{'}).Our findings highlight pitfalls in attributing human reasoning behaviours to LLMs as long as consistent reasoning remains elusive.",
}
| Recent scholarship on reasoning in LLMs has supplied evidence of impressive performance and flexible adaptation to machine generated or human critique. Nonmonotonic reasoning, crucial to human cognition for navigating the real world, remains a challenging, yet understudied task. In this work, we study nonmonotonic reasoning capabilities of seven state-of-the-art LLMs in one abstract and one commonsense reasoning task featuring generics, such as {`}Birds fly{'}, and exceptions, {`}Penguins don{'}t fly{'} (see Fig. 1). While LLMs exhibit reasoning patterns in accordance with human nonmonotonic reasoning abilities, they fail to maintain stable beliefs on truth conditions of generics at the addition of supporting examples ({`}Owls fly{'}) or unrelated information ({`}Lions have manes{'}).Our findings highlight pitfalls in attributing human reasoning behaviours to LLMs as long as consistent reasoning remains elusive. | [
"Leidinger, Alina",
"Van Rooij, Robert",
"Shutova, Ekaterina"
]
| Are {LLM}s classical or nonmonotonic reasoners? Lessons from generics | acl-short.51 | Poster | 2406.06590v2 |
https://aclanthology.org/2024.acl-short.52.bib | @inproceedings{petridis-etal-2024-constitutionalexperts,
title = "{C}onstitutional{E}xperts: Training a Mixture of Principle-based Prompts",
author = "Petridis, Savvas and
Wedin, Ben and
Yuan, Ann and
Wexler, James and
Thain, Nithum",
editor = "Ku, Lun-Wei and
Martins, Andre and
Srikumar, Vivek",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-short.52",
pages = "574--582",
abstract = "Large language models (LLMs) are highly capable at a variety of tasks given the right prompt, but writing one is still a difficult and tedious process. In this work, we introduce ConstitutionalExperts, a method for learning a prompt consisting of constitutional principles (i.e. rules), given a training dataset. Unlike prior methods that optimize the prompt as a single entity, our method incrementally improves the prompt by surgically editing individual principles. We also show that we can improve overall performance by learning unique prompts for different semantic regions of the training data and using a mixture-of-experts (MoE) architecture to route inputs at inference time. We compare our method to other state of the art prompt-optimization techniques across six benchmark datasets. We also investigate whether MoE improves these other techniques. Our results suggest that ConstitutionalExperts outperforms other prompt optimization techniques by 10.9{\%} (F1) and that mixture-of-experts improves all techniques, suggesting its broad applicability.",
}
| Large language models (LLMs) are highly capable at a variety of tasks given the right prompt, but writing one is still a difficult and tedious process. In this work, we introduce ConstitutionalExperts, a method for learning a prompt consisting of constitutional principles (i.e. rules), given a training dataset. Unlike prior methods that optimize the prompt as a single entity, our method incrementally improves the prompt by surgically editing individual principles. We also show that we can improve overall performance by learning unique prompts for different semantic regions of the training data and using a mixture-of-experts (MoE) architecture to route inputs at inference time. We compare our method to other state of the art prompt-optimization techniques across six benchmark datasets. We also investigate whether MoE improves these other techniques. Our results suggest that ConstitutionalExperts outperforms other prompt optimization techniques by 10.9{\%} (F1) and that mixture-of-experts improves all techniques, suggesting its broad applicability. | [
"Petridis, Savvas",
"Wedin, Ben",
"Yuan, Ann",
"Wexler, James",
"Thain, Nithum"
]
| {C}onstitutional{E}xperts: Training a Mixture of Principle-based Prompts | acl-short.52 | Poster | 2312.03734v1 |
https://aclanthology.org/2024.acl-short.53.bib | @inproceedings{ge-etal-2024-time,
title = "Time Sensitive Knowledge Editing through Efficient Finetuning",
author = "Ge, Xiou and
Mousavi, Ali and
Grave, Edouard and
Joulin, Armand and
Qian, Kun and
Han, Benjamin and
Arefiyan, Mostafa and
Li, Yunyao",
editor = "Ku, Lun-Wei and
Martins, Andre and
Srikumar, Vivek",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-short.53",
pages = "583--593",
abstract = "Large Language Models (LLMs) have demonstrated impressive capability in different tasks and are bringing transformative changes to many domains. However, keeping the knowledge in LLMs up-to-date remains a challenge once pretraining is complete. It is thus essential to design effective methods to both update obsolete knowledge and induce new knowledge into LLMs. Existing locate-and-edit knowledge editing (KE) method suffers from two limitations. First, the post-edit LLMs by such methods generally have poor capability in answering complex queries that require multi-hop reasoning. Second, the long run-time of such locate-and-edit methods to perform knowledge edits make it infeasible for large scale KE in practice. In this paper, we explore Parameter-Efficient Fine-Tuning (PEFT) techniques as an alternative for KE. We curate a more comprehensive temporal KE dataset with both knowledge update and knowledge injection examples for KE performance benchmarking. We further probe the effect of fine-tuning on a range of layers in an LLM for the multi-hop QA task. We find that PEFT performs better than locate-and-edit techniques for time-sensitive knowledge edits.",
}
| Large Language Models (LLMs) have demonstrated impressive capability in different tasks and are bringing transformative changes to many domains. However, keeping the knowledge in LLMs up-to-date remains a challenge once pretraining is complete. It is thus essential to design effective methods to both update obsolete knowledge and induce new knowledge into LLMs. Existing locate-and-edit knowledge editing (KE) method suffers from two limitations. First, the post-edit LLMs by such methods generally have poor capability in answering complex queries that require multi-hop reasoning. Second, the long run-time of such locate-and-edit methods to perform knowledge edits make it infeasible for large scale KE in practice. In this paper, we explore Parameter-Efficient Fine-Tuning (PEFT) techniques as an alternative for KE. We curate a more comprehensive temporal KE dataset with both knowledge update and knowledge injection examples for KE performance benchmarking. We further probe the effect of fine-tuning on a range of layers in an LLM for the multi-hop QA task. We find that PEFT performs better than locate-and-edit techniques for time-sensitive knowledge edits. | [
"Ge, Xiou",
"Mousavi, Ali",
"Grave, Edouard",
"Joulin, Arm",
"",
"Qian, Kun",
"Han, Benjamin",
"Arefiyan, Mostafa",
"Li, Yunyao"
]
| Time Sensitive Knowledge Editing through Efficient Finetuning | acl-short.53 | Poster | 2406.04496v2 |
https://aclanthology.org/2024.acl-short.54.bib | @inproceedings{kong-etal-2024-prewrite,
title = "{PR}ewrite: Prompt Rewriting with Reinforcement Learning",
author = "Kong, Weize and
Hombaiah, Spurthi and
Zhang, Mingyang and
Mei, Qiaozhu and
Bendersky, Michael",
editor = "Ku, Lun-Wei and
Martins, Andre and
Srikumar, Vivek",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-short.54",
pages = "594--601",
abstract = "Prompt engineering is critical for the development of LLM-based applications. However, it is usually done manually in a {``}trial and error{''} fashion that can be time consuming, ineffective, and sub-optimal. Even for the prompts which seemingly work well, there is always a lingering question: can the prompts be made better with further modifications?To address these problems, we investigate automated prompt engineering in this paper. Specifically, we propose PRewrite, an automated method to rewrite an under-optimized prompt to a more effective prompt. We instantiate the prompt rewriter using an LLM. The rewriter LLM is trained using reinforcement learning to optimize the performance on a given downstream task. We conduct experiments on diverse benchmark datasets, which demonstrates the effectiveness of PRewrite.",
}
| Prompt engineering is critical for the development of LLM-based applications. However, it is usually done manually in a {``}trial and error{''} fashion that can be time consuming, ineffective, and sub-optimal. Even for the prompts which seemingly work well, there is always a lingering question: can the prompts be made better with further modifications?To address these problems, we investigate automated prompt engineering in this paper. Specifically, we propose PRewrite, an automated method to rewrite an under-optimized prompt to a more effective prompt. We instantiate the prompt rewriter using an LLM. The rewriter LLM is trained using reinforcement learning to optimize the performance on a given downstream task. We conduct experiments on diverse benchmark datasets, which demonstrates the effectiveness of PRewrite. | [
"Kong, Weize",
"Hombaiah, Spurthi",
"Zhang, Mingyang",
"Mei, Qiaozhu",
"Bendersky, Michael"
]
| {PR}ewrite: Prompt Rewriting with Reinforcement Learning | acl-short.54 | Poster | 2407.12794v1 |
https://aclanthology.org/2024.acl-short.55.bib | @inproceedings{rezaei-blanco-2024-paraphrasing,
title = "Paraphrasing in Affirmative Terms Improves Negation Understanding",
author = "Rezaei, MohammadHossein and
Blanco, Eduardo",
editor = "Ku, Lun-Wei and
Martins, Andre and
Srikumar, Vivek",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-short.55",
pages = "602--615",
abstract = "Negation is a common linguistic phenomenon. Yet language models face challenges with negation in many natural language understanding tasks such as question answering and natural language inference. In this paper, we experiment with seamless strategies that incorporate affirmative interpretations (i.e., paraphrases without negation) to make models more robust against negation. Crucially, our affirmative interpretations are obtained automatically. We show improvements with CondaQA, a large corpus requiring reasoning with negation, and five natural language understanding tasks.",
}
| Negation is a common linguistic phenomenon. Yet language models face challenges with negation in many natural language understanding tasks such as question answering and natural language inference. In this paper, we experiment with seamless strategies that incorporate affirmative interpretations (i.e., paraphrases without negation) to make models more robust against negation. Crucially, our affirmative interpretations are obtained automatically. We show improvements with CondaQA, a large corpus requiring reasoning with negation, and five natural language understanding tasks. | [
"Rezaei, MohammadHossein",
"Blanco, Eduardo"
]
| Paraphrasing in Affirmative Terms Improves Negation Understanding | acl-short.55 | Poster | 2406.07492v1 |
https://aclanthology.org/2024.acl-short.56.bib | @inproceedings{sun-etal-2024-exploring-conditional,
title = "Exploring Conditional Variational Mechanism to {P}inyin Input Method for Addressing One-to-Many Mappings in Low-Resource Scenarios",
author = "Sun, Bin and
Li, Jianfeng and
Zhou, Hao and
Meng, Fandong and
Li, Kan and
Zhou, Jie",
editor = "Ku, Lun-Wei and
Martins, Andre and
Srikumar, Vivek",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-short.56",
pages = "616--629",
abstract = "Pinyin input method engine (IME) refers to the transformation tool from pinyin sequence to Chinese characters, which is widely used on mobile phone applications. Due to the homophones, Pinyin IME suffers from the one-to-many mapping problem in the process of pinyin sequences to Chinese characters. To solve the above issue, this paper makes the first exploration to leverage an effective conditional variational mechanism (CVM) for pinyin IME. However, to ensure the stable and smooth operation of Pinyin IME under low-resource conditions (e.g., on offline mobile devices), we should balance diversity, accuracy, and efficiency with CVM, which is still challenging. To this end, we employ a novel strategy that simplifies the complexity of semantic encoding by facilitating the interaction between pinyin and the Chinese character information during the construction of continuous latent variables. Concurrently, the accuracy of the outcomes is enhanced by capitalizing on the discrete latent variables. Experimental results demonstrate the superior performance of our method.",
}
| Pinyin input method engine (IME) refers to the transformation tool from pinyin sequence to Chinese characters, which is widely used on mobile phone applications. Due to the homophones, Pinyin IME suffers from the one-to-many mapping problem in the process of pinyin sequences to Chinese characters. To solve the above issue, this paper makes the first exploration to leverage an effective conditional variational mechanism (CVM) for pinyin IME. However, to ensure the stable and smooth operation of Pinyin IME under low-resource conditions (e.g., on offline mobile devices), we should balance diversity, accuracy, and efficiency with CVM, which is still challenging. To this end, we employ a novel strategy that simplifies the complexity of semantic encoding by facilitating the interaction between pinyin and the Chinese character information during the construction of continuous latent variables. Concurrently, the accuracy of the outcomes is enhanced by capitalizing on the discrete latent variables. Experimental results demonstrate the superior performance of our method. | [
"Sun, Bin",
"Li, Jianfeng",
"Zhou, Hao",
"Meng, F",
"ong",
"Li, Kan",
"Zhou, Jie"
]
| Exploring Conditional Variational Mechanism to {P}inyin Input Method for Addressing One-to-Many Mappings in Low-Resource Scenarios | acl-short.56 | Poster | 2203.00249v2 |
https://aclanthology.org/2024.acl-short.57.bib | @inproceedings{hemati-beigy-2024-consistency,
title = "Consistency Training by Synthetic Question Generation for Conversational Question Answering",
author = "Hemati, Hamed and
Beigy, Hamid",
editor = "Ku, Lun-Wei and
Martins, Andre and
Srikumar, Vivek",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-short.57",
pages = "630--639",
abstract = "Efficiently modeling historical information is a critical component in addressing user queries within a conversational question-answering (QA) context, as historical context plays a vital role in clarifying the user{'}s questions. However, irrelevant history induces noise in the reasoning process, especially for those questions with a considerable historical context. In our novel model-agnostic approach, referred to as **CoTaH** (**Co**nsistency-**T**rained **a**ugmented **H**istory), we augment the historical information with synthetic questions and subsequently employ consistency training to train a model that utilizes both real and augmented historical data to implicitly make the reasoning robust to irrelevant history. To the best of our knowledge, this is the first instance of research using synthetic question generation as a form of data augmentation to model conversational QA settings. By citing a common modeling error prevalent in previous research, we introduce a new baseline and compare our model{'}s performance against it, demonstrating an improvement in results, particularly in later turns of the conversation, when dealing with questions that include a large historical context.",
}
| Efficiently modeling historical information is a critical component in addressing user queries within a conversational question-answering (QA) context, as historical context plays a vital role in clarifying the user{'}s questions. However, irrelevant history induces noise in the reasoning process, especially for those questions with a considerable historical context. In our novel model-agnostic approach, referred to as **CoTaH** (**Co**nsistency-**T**rained **a**ugmented **H**istory), we augment the historical information with synthetic questions and subsequently employ consistency training to train a model that utilizes both real and augmented historical data to implicitly make the reasoning robust to irrelevant history. To the best of our knowledge, this is the first instance of research using synthetic question generation as a form of data augmentation to model conversational QA settings. By citing a common modeling error prevalent in previous research, we introduce a new baseline and compare our model{'}s performance against it, demonstrating an improvement in results, particularly in later turns of the conversation, when dealing with questions that include a large historical context. | [
"Hemati, Hamed",
"Beigy, Hamid"
]
| Consistency Training by Synthetic Question Generation for Conversational Question Answering | acl-short.57 | Poster | 2404.11109v1 |
https://aclanthology.org/2024.acl-short.58.bib | @inproceedings{singh-etal-2024-good,
title = "How Good is Zero-Shot {MT} Evaluation for Low Resource {I}ndian Languages?",
author = "Singh, Anushka and
Sai, Ananya and
Dabre, Raj and
Puduppully, Ratish and
Kunchukuttan, Anoop and
Khapra, Mitesh",
editor = "Ku, Lun-Wei and
Martins, Andre and
Srikumar, Vivek",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-short.58",
pages = "640--649",
abstract = "While machine translation evaluation has been studied primarily for high-resource languages, there has been a recent interest in evaluation for low-resource languages due to the increasing availability of data and models. In this paper, we focus on a zero-shot evaluation setting focusing on low-resource Indian languages, namely Assamese, Kannada, Maithili, and Punjabi. We collect sufficient Multi-Dimensional Quality Metrics (MQM) and Direct Assessment (DA) annotations to create test sets and meta-evaluate a plethora of automatic evaluation metrics. We observe that even for learned metrics, which are known to exhibit zero-shot performance, the Kendall Tau and Pearson correlations with human annotations are only as high as 0.32 and 0.45. Synthetic data approaches show mixed results and overall do not help close the gap by much for these languages. This indicates that there is still a long way to go for low-resource evaluation.",
}
| While machine translation evaluation has been studied primarily for high-resource languages, there has been a recent interest in evaluation for low-resource languages due to the increasing availability of data and models. In this paper, we focus on a zero-shot evaluation setting focusing on low-resource Indian languages, namely Assamese, Kannada, Maithili, and Punjabi. We collect sufficient Multi-Dimensional Quality Metrics (MQM) and Direct Assessment (DA) annotations to create test sets and meta-evaluate a plethora of automatic evaluation metrics. We observe that even for learned metrics, which are known to exhibit zero-shot performance, the Kendall Tau and Pearson correlations with human annotations are only as high as 0.32 and 0.45. Synthetic data approaches show mixed results and overall do not help close the gap by much for these languages. This indicates that there is still a long way to go for low-resource evaluation. | [
"Singh, Anushka",
"Sai, Ananya",
"Dabre, Raj",
"Puduppully, Ratish",
"Kunchukuttan, Anoop",
"Khapra, Mitesh"
]
| How Good is Zero-Shot {MT} Evaluation for Low Resource {I}ndian Languages? | acl-short.58 | Poster | 2310.09765v1 |
https://aclanthology.org/2024.acl-short.59.bib | @inproceedings{adeyemi-etal-2024-zero,
title = "Zero-Shot Cross-Lingual Reranking with Large Language Models for Low-Resource Languages",
author = "Adeyemi, Mofetoluwa and
Oladipo, Akintunde and
Pradeep, Ronak and
Lin, Jimmy",
editor = "Ku, Lun-Wei and
Martins, Andre and
Srikumar, Vivek",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-short.59",
pages = "650--656",
abstract = "Large language models (LLMs) as listwise rerankers have shown impressive zero-shot capabilities in various passage ranking tasks. Despite their success, there is still a gap in existing literature on their effectiveness in reranking low-resource languages. To address this, we investigate how LLMs function as listwise rerankers in cross-lingual information retrieval (CLIR) systems with queries in English and passages in four African languages: Hausa, Somali, Swahili, and Yoruba. We analyze and compare the effectiveness of monolingual reranking using either query or document translations. We also evaluate the effectiveness of LLMs when leveraging their own generated translations. To grasp the general picture, we examine the effectiveness of multiple LLMs {---} the proprietary models RankGPT-4 and RankGPT-3.5, along with the open-source model RankZephyr. While the document translation setting, i.e., both queries and documents are in English, leads to the best reranking effectiveness, our results indicate that for specific LLMs, reranking in the African language setting achieves competitive effectiveness with the cross-lingual setting, and even performs better when using the LLM{'}s own translations.",
}
| Large language models (LLMs) as listwise rerankers have shown impressive zero-shot capabilities in various passage ranking tasks. Despite their success, there is still a gap in existing literature on their effectiveness in reranking low-resource languages. To address this, we investigate how LLMs function as listwise rerankers in cross-lingual information retrieval (CLIR) systems with queries in English and passages in four African languages: Hausa, Somali, Swahili, and Yoruba. We analyze and compare the effectiveness of monolingual reranking using either query or document translations. We also evaluate the effectiveness of LLMs when leveraging their own generated translations. To grasp the general picture, we examine the effectiveness of multiple LLMs {---} the proprietary models RankGPT-4 and RankGPT-3.5, along with the open-source model RankZephyr. While the document translation setting, i.e., both queries and documents are in English, leads to the best reranking effectiveness, our results indicate that for specific LLMs, reranking in the African language setting achieves competitive effectiveness with the cross-lingual setting, and even performs better when using the LLM{'}s own translations. | [
"Adeyemi, Mofetoluwa",
"Oladipo, Akintunde",
"Pradeep, Ronak",
"Lin, Jimmy"
]
| Zero-Shot Cross-Lingual Reranking with Large Language Models for Low-Resource Languages | acl-short.59 | Poster | 2312.16159v1 |
https://aclanthology.org/2024.acl-short.60.bib | @inproceedings{verma-etal-2024-cross,
title = "Cross-Modal Projection in Multimodal {LLM}s Doesn{'}t Really Project Visual Attributes to Textual Space",
author = "Verma, Gaurav and
Choi, Minje and
Sharma, Kartik and
Watson-Daniels, Jamelle and
Oh, Sejoon and
Kumar, Srijan",
editor = "Ku, Lun-Wei and
Martins, Andre and
Srikumar, Vivek",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-short.60",
pages = "657--664",
abstract = "Multimodal large language models (MLLMs) like LLaVA and GPT-4(V) enable general-purpose conversations about images with the language modality. As off-the-shelf MLLMs may have limited capabilities on images from domains like dermatology and agriculture, they must be fine-tuned to unlock domain-specific applications. The prevalent architecture of current open-source MLLMs comprises two major modules: an image-language (cross-modal) projection network and a large language model. It is desirable to understand the roles of these two modules in modeling domain-specific visual attributes to inform the design of future models and streamline the interpretability efforts on the current models. To this end, via experiments on 4 datasets and under 2 fine-tuning settings, we find that as the MLLM is fine-tuned, it indeed gains domain-specific visual capabilities, but the updates do not lead to the projection extracting relevant domain-specific visual attributes. Our results indicate that the domain-specific visual attributes are modeled by the LLM, even when only the projection is fine-tuned. Through this study, we offer a potential reinterpretation of the role of cross-modal projections in MLLM architectures.",
}
| Multimodal large language models (MLLMs) like LLaVA and GPT-4(V) enable general-purpose conversations about images with the language modality. As off-the-shelf MLLMs may have limited capabilities on images from domains like dermatology and agriculture, they must be fine-tuned to unlock domain-specific applications. The prevalent architecture of current open-source MLLMs comprises two major modules: an image-language (cross-modal) projection network and a large language model. It is desirable to understand the roles of these two modules in modeling domain-specific visual attributes to inform the design of future models and streamline the interpretability efforts on the current models. To this end, via experiments on 4 datasets and under 2 fine-tuning settings, we find that as the MLLM is fine-tuned, it indeed gains domain-specific visual capabilities, but the updates do not lead to the projection extracting relevant domain-specific visual attributes. Our results indicate that the domain-specific visual attributes are modeled by the LLM, even when only the projection is fine-tuned. Through this study, we offer a potential reinterpretation of the role of cross-modal projections in MLLM architectures. | [
"Verma, Gaurav",
"Choi, Minje",
"Sharma, Kartik",
"Watson-Daniels, Jamelle",
"Oh, Sejoon",
"Kumar, Srijan"
]
| Cross-Modal Projection in Multimodal {LLM}s Doesn{'}t Really Project Visual Attributes to Textual Space | acl-short.60 | Poster | 2402.16832v2 |
https://aclanthology.org/2024.acl-short.61.bib | @inproceedings{kang-etal-2024-guidance,
title = "Guidance-Based Prompt Data Augmentation in Specialized Domains for Named Entity Recognition",
author = "Kang, Hyeonseok and
Seo, Hyein and
Jung, Jeesu and
Jung, Sangkeun and
Chang, Du-Seong and
Chung, Riwoo",
editor = "Ku, Lun-Wei and
Martins, Andre and
Srikumar, Vivek",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-short.61",
pages = "665--672",
abstract = "While the abundance of rich and vast datasets across numerous fields has facilitated the advancement of natural language processing, sectors in need of specialized data types continue to struggle with the challenge of finding quality data. Our study introduces a novel guidance data augmentation technique utilizing abstracted context and sentence structures to produce varied sentences while maintaining context-entity relationships, addressing data scarcity challenges. By fostering a closer relationship between context, sentence structure, and role of entities, our method enhances data augmentation{'}s effectiveness. Consequently, by showcasing diversification in both entity-related vocabulary and overall sentence structure, and simultaneously improving the training performance of named entity recognition task.",
}
| While the abundance of rich and vast datasets across numerous fields has facilitated the advancement of natural language processing, sectors in need of specialized data types continue to struggle with the challenge of finding quality data. Our study introduces a novel guidance data augmentation technique utilizing abstracted context and sentence structures to produce varied sentences while maintaining context-entity relationships, addressing data scarcity challenges. By fostering a closer relationship between context, sentence structure, and role of entities, our method enhances data augmentation{'}s effectiveness. Consequently, by showcasing diversification in both entity-related vocabulary and overall sentence structure, and simultaneously improving the training performance of named entity recognition task. | [
"Kang, Hyeonseok",
"Seo, Hyein",
"Jung, Jeesu",
"Jung, Sangkeun",
"Chang, Du-Seong",
"Chung, Riwoo"
]
| Guidance-Based Prompt Data Augmentation in Specialized Domains for Named Entity Recognition | acl-short.61 | Poster | 2407.18442v1 |
https://aclanthology.org/2024.acl-short.62.bib | @inproceedings{xu-etal-2024-aligning,
title = "Aligning Large Language Models via Fine-grained Supervision",
author = "Xu, Dehong and
Qiu, Liang and
Kim, Minseok and
Ladhak, Faisal and
Do, Jaeyoung",
editor = "Ku, Lun-Wei and
Martins, Andre and
Srikumar, Vivek",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-short.62",
pages = "673--680",
abstract = "Pre-trained large-scale language models (LLMs) excel at producing coherent articles, yet their outputs may be untruthful, toxic, or fail to align with user expectations. Current approaches focus on using reinforcement learning with human feedback (RLHF) to improve model alignment, which works by transforming coarse human preferences of LLM outputs into a feedback signal that guides the model learning process. However, because this approach operates on sequence-level feedback, it lacks the precision to identify the exact parts of the output affecting user preferences. To address this gap, we propose a method to enhance LLM alignment through fine-grained token-level supervision. Specifically, we ask annotators to minimally edit less preferred responses within the standard reward modeling dataset to make them more favorable, ensuring changes are made only where necessary while retaining most of the original content. The refined dataset is used to train a token-level reward model, which is then used for training our fine-grained Proximal Policy Optimization (PPO) model. Our experiment results demonstrate that this approach can improve LLM performance by up to 5.1{\%} in terms of win rate against the reference model, compared with the traditional PPO model.",
}
| Pre-trained large-scale language models (LLMs) excel at producing coherent articles, yet their outputs may be untruthful, toxic, or fail to align with user expectations. Current approaches focus on using reinforcement learning with human feedback (RLHF) to improve model alignment, which works by transforming coarse human preferences of LLM outputs into a feedback signal that guides the model learning process. However, because this approach operates on sequence-level feedback, it lacks the precision to identify the exact parts of the output affecting user preferences. To address this gap, we propose a method to enhance LLM alignment through fine-grained token-level supervision. Specifically, we ask annotators to minimally edit less preferred responses within the standard reward modeling dataset to make them more favorable, ensuring changes are made only where necessary while retaining most of the original content. The refined dataset is used to train a token-level reward model, which is then used for training our fine-grained Proximal Policy Optimization (PPO) model. Our experiment results demonstrate that this approach can improve LLM performance by up to 5.1{\%} in terms of win rate against the reference model, compared with the traditional PPO model. | [
"Xu, Dehong",
"Qiu, Liang",
"Kim, Minseok",
"Ladhak, Faisal",
"Do, Jaeyoung"
]
| Aligning Large Language Models via Fine-grained Supervision | acl-short.62 | Poster | 2010.06775v1 |
https://aclanthology.org/2024.acl-short.63.bib | @inproceedings{cui-swayamdipta-2024-annotating,
title = "Annotating {F}rame{N}et via Structure-Conditioned Language Generation",
author = "Cui, Xinyue and
Swayamdipta, Swabha",
editor = "Ku, Lun-Wei and
Martins, Andre and
Srikumar, Vivek",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-short.63",
pages = "681--692",
abstract = "Despite the remarkable generative capabilities of language models in producing naturalistic language, their effectiveness on explicit manipulation and generation of linguistic structures remain understudied. In this paper, we investigate the task of generating new sentences preserving a given semantic structure, following the FrameNet formalism. We propose a framework to produce novel frame-semantically annotated sentences following an overgenerate-and-filter approach. Our results show that conditioning on rich, explicit semantic information tends to produce generations with high human acceptance, under both prompting and finetuning. Our generated frame-semantic structured annotations are effective at training data augmentation for frame-semantic role labeling in low-resource settings; however, we do not see benefits under higher resource settings. Our study concludes that while generating high-quality, semantically rich data might be within reach, the downstream utility of such generations remains to be seen, highlighting the outstanding challenges with automating linguistic annotation tasks.",
}
| Despite the remarkable generative capabilities of language models in producing naturalistic language, their effectiveness on explicit manipulation and generation of linguistic structures remain understudied. In this paper, we investigate the task of generating new sentences preserving a given semantic structure, following the FrameNet formalism. We propose a framework to produce novel frame-semantically annotated sentences following an overgenerate-and-filter approach. Our results show that conditioning on rich, explicit semantic information tends to produce generations with high human acceptance, under both prompting and finetuning. Our generated frame-semantic structured annotations are effective at training data augmentation for frame-semantic role labeling in low-resource settings; however, we do not see benefits under higher resource settings. Our study concludes that while generating high-quality, semantically rich data might be within reach, the downstream utility of such generations remains to be seen, highlighting the outstanding challenges with automating linguistic annotation tasks. | [
"Cui, Xinyue",
"Swayamdipta, Swabha"
]
| Annotating {F}rame{N}et via Structure-Conditioned Language Generation | acl-short.63 | Poster | 1511.03924v1 |
https://aclanthology.org/2024.acl-short.64.bib | @inproceedings{chen-etal-2024-dual,
title = "{DUAL}-{REFLECT}: Enhancing Large Language Models for Reflective Translation through Dual Learning Feedback Mechanisms",
author = "Chen, Andong and
Lou, Lianzhang and
Chen, Kehai and
Bai, Xuefeng and
Xiang, Yang and
Yang, Muyun and
Zhao, Tiejun and
Zhang, Min",
editor = "Ku, Lun-Wei and
Martins, Andre and
Srikumar, Vivek",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-short.64",
pages = "693--704",
abstract = "Recently, large language models (LLMs) enhanced by self-reflection have achieved promising performance on machine transla004 tion. The key idea is guiding LLMs to generate translation with human-like feedback. However, existing self-reflection methods lack effective feedback information, limiting the translation performance. To address this, we introduce a DUAL-REFLECT framework, leveraging the dual learning of translation tasks to provide effective feedback, thereby enhancing the models{'} self-reflective abilities and improving translation performance. The application of this method across various translation tasks has proven its effectiveness in improving translation accuracy and eliminating ambiguities, especially in translation tasks with low-resource language pairs.",
}
| Recently, large language models (LLMs) enhanced by self-reflection have achieved promising performance on machine transla004 tion. The key idea is guiding LLMs to generate translation with human-like feedback. However, existing self-reflection methods lack effective feedback information, limiting the translation performance. To address this, we introduce a DUAL-REFLECT framework, leveraging the dual learning of translation tasks to provide effective feedback, thereby enhancing the models{'} self-reflective abilities and improving translation performance. The application of this method across various translation tasks has proven its effectiveness in improving translation accuracy and eliminating ambiguities, especially in translation tasks with low-resource language pairs. | [
"Chen, Andong",
"Lou, Lianzhang",
"Chen, Kehai",
"Bai, Xuefeng",
"Xiang, Yang",
"Yang, Muyun",
"Zhao, Tiejun",
"Zhang, Min"
]
| {DUAL}-{REFLECT}: Enhancing Large Language Models for Reflective Translation through Dual Learning Feedback Mechanisms | acl-short.64 | Poster | 2406.07232v2 |
https://aclanthology.org/2024.acl-short.65.bib | @inproceedings{hayashi-etal-2024-towards,
title = "Towards Artwork Explanation in Large-scale Vision Language Models",
author = "Hayashi, Kazuki and
Sakai, Yusuke and
Kamigaito, Hidetaka and
Hayashi, Katsuhiko and
Watanabe, Taro",
editor = "Ku, Lun-Wei and
Martins, Andre and
Srikumar, Vivek",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-short.65",
pages = "705--729",
abstract = "Large-scale Vision-Language Models (LVLMs) output text from images and instructions, demonstrating advanced capabilities in text generation and comprehension. However, it has not been clarified to what extent LVLMs understand the knowledge necessary for explaining images, the complex relationships between various pieces of knowledge, and how they integrate these understandings into their explanations. To address this issue, we propose a new task: the artwork explanation generation task, along with its evaluation dataset and metric for quantitatively assessing the understanding and utilization of knowledge about artworks. This task is apt for image description based on the premise that LVLMs are expected to have pre-existing knowledge of artworks, which are often subjects of wide recognition and documented information.It consists of two parts: generating explanations from both images and titles of artworks, and generating explanations using only images, thus evaluating the LVLMs{'} language-based and vision-based knowledge.Alongside, we release a training dataset for LVLMs to learn explanations that incorporate knowledge about artworks.Our findings indicate that LVLMs not only struggle with integrating language and visual information but also exhibit a more pronounced limitation in acquiring knowledge from images alone. The datasets ExpArt=Explain Artworks are available at https://huggingface.co/datasets/naist-nlp/ExpArt",
}
| Large-scale Vision-Language Models (LVLMs) output text from images and instructions, demonstrating advanced capabilities in text generation and comprehension. However, it has not been clarified to what extent LVLMs understand the knowledge necessary for explaining images, the complex relationships between various pieces of knowledge, and how they integrate these understandings into their explanations. To address this issue, we propose a new task: the artwork explanation generation task, along with its evaluation dataset and metric for quantitatively assessing the understanding and utilization of knowledge about artworks. This task is apt for image description based on the premise that LVLMs are expected to have pre-existing knowledge of artworks, which are often subjects of wide recognition and documented information.It consists of two parts: generating explanations from both images and titles of artworks, and generating explanations using only images, thus evaluating the LVLMs{'} language-based and vision-based knowledge.Alongside, we release a training dataset for LVLMs to learn explanations that incorporate knowledge about artworks.Our findings indicate that LVLMs not only struggle with integrating language and visual information but also exhibit a more pronounced limitation in acquiring knowledge from images alone. The datasets ExpArt=Explain Artworks are available at https://huggingface.co/datasets/naist-nlp/ExpArt | [
"Hayashi, Kazuki",
"Sakai, Yusuke",
"Kamigaito, Hidetaka",
"Hayashi, Katsuhiko",
"Watanabe, Taro"
]
| Towards Artwork Explanation in Large-scale Vision Language Models | acl-short.65 | Poster | 2403.00068v1 |
https://aclanthology.org/2024.acl-short.66.bib | @inproceedings{zhong-etal-2024-hallucination,
title = "On the Hallucination in Simultaneous Machine Translation",
author = "Zhong, Meizhi and
Chen, Kehai and
Xue, Zhengshan and
Liu, Lemao and
Yang, Mingming and
Zhang, Min",
editor = "Ku, Lun-Wei and
Martins, Andre and
Srikumar, Vivek",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-short.66",
pages = "730--742",
abstract = "It is widely known that hallucination is a critical issue in Simultaneous Machine Translation (SiMT) due to the absence of source-side information. While many efforts have been made to enhance performance for SiMT, few of them attempt to understand and analyze hallucination in SiMT.Therefore, we conduct a comprehensive analysis of hallucination in SiMT from two perspectives: understanding the distribution of hallucination words and the target-side context usage of them.Intensive experiments demonstrate some valuable findings and particularly show that it is possible to alleviate hallucination by decreasing the over usage of target-side information for SiMT.",
}
| It is widely known that hallucination is a critical issue in Simultaneous Machine Translation (SiMT) due to the absence of source-side information. While many efforts have been made to enhance performance for SiMT, few of them attempt to understand and analyze hallucination in SiMT.Therefore, we conduct a comprehensive analysis of hallucination in SiMT from two perspectives: understanding the distribution of hallucination words and the target-side context usage of them.Intensive experiments demonstrate some valuable findings and particularly show that it is possible to alleviate hallucination by decreasing the over usage of target-side information for SiMT. | [
"Zhong, Meizhi",
"Chen, Kehai",
"Xue, Zhengshan",
"Liu, Lemao",
"Yang, Mingming",
"Zhang, Min"
]
| On the Hallucination in Simultaneous Machine Translation | acl-short.66 | Poster | 2406.07239v1 |
https://aclanthology.org/2024.acl-short.67.bib | @inproceedings{li-etal-2024-self-augmented,
title = "Self-Augmented In-Context Learning for Unsupervised Word Translation",
author = "Li, Yaoyiran and
Korhonen, Anna and
Vuli{\'c}, Ivan",
editor = "Ku, Lun-Wei and
Martins, Andre and
Srikumar, Vivek",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-short.67",
pages = "743--753",
abstract = "Recent work has shown that, while large language models (LLMs) demonstrate strong word translation or bilingual lexicon induction (BLI) capabilities in few-shot setups, they still cannot match the performance of {`}traditional{'} mapping-based approaches in the unsupervised scenario where no seed translation pairs are available, especially for lower-resource languages. To address this challenge with LLMs, we propose self-augmented in-context learning (SAIL) for unsupervised BLI: starting from a zero-shot prompt, SAIL iteratively induces a set of high-confidence word translation pairs for in-context learning (ICL) from an LLM, which it then reapplies to the same LLM in the ICL fashion. Our method shows substantial gains over zero-shot prompting of LLMs on two established BLI benchmarks spanning a wide range of language pairs, also outperforming mapping-based baselines across the board. In addition to achieving state-of-the-art unsupervised BLI performance, we also conduct comprehensive analyses on SAIL and discuss its limitations.",
}
| Recent work has shown that, while large language models (LLMs) demonstrate strong word translation or bilingual lexicon induction (BLI) capabilities in few-shot setups, they still cannot match the performance of {`}traditional{'} mapping-based approaches in the unsupervised scenario where no seed translation pairs are available, especially for lower-resource languages. To address this challenge with LLMs, we propose self-augmented in-context learning (SAIL) for unsupervised BLI: starting from a zero-shot prompt, SAIL iteratively induces a set of high-confidence word translation pairs for in-context learning (ICL) from an LLM, which it then reapplies to the same LLM in the ICL fashion. Our method shows substantial gains over zero-shot prompting of LLMs on two established BLI benchmarks spanning a wide range of language pairs, also outperforming mapping-based baselines across the board. In addition to achieving state-of-the-art unsupervised BLI performance, we also conduct comprehensive analyses on SAIL and discuss its limitations. | [
"Li, Yaoyiran",
"Korhonen, Anna",
"Vuli{\\'c}, Ivan"
]
| Self-Augmented In-Context Learning for Unsupervised Word Translation | acl-short.67 | Poster | 2006.11578v1 |
https://aclanthology.org/2024.acl-short.68.bib | @inproceedings{xu-etal-2024-ram,
title = "{RAM}-{EHR}: Retrieval Augmentation Meets Clinical Predictions on Electronic Health Records",
author = "Xu, Ran and
Shi, Wenqi and
Yu, Yue and
Zhuang, Yuchen and
Jin, Bowen and
Wang, May Dongmei and
Ho, Joyce and
Yang, Carl",
editor = "Ku, Lun-Wei and
Martins, Andre and
Srikumar, Vivek",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-short.68",
pages = "754--765",
abstract = "We present RAM-EHR, a Retrieval AugMentation pipeline to improve clinical predictions on Electronic Health Records (EHRs). RAM-EHR first collects multiple knowledge sources, converts them into text format, and uses dense retrieval to obtain information related to medical concepts. This strategy addresses the difficulties associated with complex names for the concepts. RAM-EHR then augments the local EHR predictive model co-trained with consistency regularization to capture complementary information from patient visits and summarized knowledge. Experiments on two EHR datasets show the efficacy of RAM-EHR over previous knowledge-enhanced baselines (3.4{\%} gain in AUROC and 7.2{\%} gain in AUPR), emphasizing the effectiveness of the summarized knowledge from RAM-EHR for clinical prediction tasks.",
}
| We present RAM-EHR, a Retrieval AugMentation pipeline to improve clinical predictions on Electronic Health Records (EHRs). RAM-EHR first collects multiple knowledge sources, converts them into text format, and uses dense retrieval to obtain information related to medical concepts. This strategy addresses the difficulties associated with complex names for the concepts. RAM-EHR then augments the local EHR predictive model co-trained with consistency regularization to capture complementary information from patient visits and summarized knowledge. Experiments on two EHR datasets show the efficacy of RAM-EHR over previous knowledge-enhanced baselines (3.4{\%} gain in AUROC and 7.2{\%} gain in AUPR), emphasizing the effectiveness of the summarized knowledge from RAM-EHR for clinical prediction tasks. | [
"Xu, Ran",
"Shi, Wenqi",
"Yu, Yue",
"Zhuang, Yuchen",
"Jin, Bowen",
"Wang, May Dongmei",
"Ho, Joyce",
"Yang, Carl"
]
| {RAM}-{EHR}: Retrieval Augmentation Meets Clinical Predictions on Electronic Health Records | acl-short.68 | Oral | 2403.00815v3 |
https://aclanthology.org/2024.acl-demos.1.bib | @inproceedings{wang-etal-2024-pai,
title = "{PAI}-Diffusion: Constructing and Serving a Family of Open {C}hinese Diffusion Models for Text-to-image Synthesis on the Cloud",
author = "Wang, Chengyu and
Duan, Zhongjie and
Liu, Bingyan and
Zou, Xinyi and
Chen, Cen and
Jia, Kui and
Huang, Jun",
editor = "Cao, Yixin and
Feng, Yang and
Xiong, Deyi",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 3: System Demonstrations)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-demos.1",
pages = "1--8",
abstract = "Text-to-image synthesis for the Chinese language poses unique challenges due to its large vocabulary size, and intricate character relationships. While existing diffusion models have shown promise in generating images from textual descriptions, they often neglect domain-specific contexts and lack robustness in handling the Chinese language. This paper introduces PAI-Diffusion, a comprehensive framework that addresses these limitations. PAI-Diffusion incorporates both general and domain-specific Chinese diffusion models, enabling the generation of contextually relevant images. It explores the potential of using LoRA and ControlNet for fine-grained image style transfer and image editing, empowering users with enhanced control over image generation. Moreover, PAI-Diffusion seamlessly integrates with Alibaba Cloud{'}s Platform for AI, providing accessible and scalable solutions. All the Chinese diffusion model checkpoints, LoRAs, and ControlNets, including domain-specific ones, are publicly available. A user-friendly Chinese WebUI and the diffusers-api elastic inference toolkit, also open-sourced, further facilitate the easy deployment of PAI-Diffusion models in various local and cloud environments, making it a valuable resource for Chinese text-to-image synthesis.",
}
| Text-to-image synthesis for the Chinese language poses unique challenges due to its large vocabulary size, and intricate character relationships. While existing diffusion models have shown promise in generating images from textual descriptions, they often neglect domain-specific contexts and lack robustness in handling the Chinese language. This paper introduces PAI-Diffusion, a comprehensive framework that addresses these limitations. PAI-Diffusion incorporates both general and domain-specific Chinese diffusion models, enabling the generation of contextually relevant images. It explores the potential of using LoRA and ControlNet for fine-grained image style transfer and image editing, empowering users with enhanced control over image generation. Moreover, PAI-Diffusion seamlessly integrates with Alibaba Cloud{'}s Platform for AI, providing accessible and scalable solutions. All the Chinese diffusion model checkpoints, LoRAs, and ControlNets, including domain-specific ones, are publicly available. A user-friendly Chinese WebUI and the diffusers-api elastic inference toolkit, also open-sourced, further facilitate the easy deployment of PAI-Diffusion models in various local and cloud environments, making it a valuable resource for Chinese text-to-image synthesis. | [
"Wang, Chengyu",
"Duan, Zhongjie",
"Liu, Bingyan",
"Zou, Xinyi",
"Chen, Cen",
"Jia, Kui",
"Huang, Jun"
]
| {PAI}-Diffusion: Constructing and Serving a Family of Open {C}hinese Diffusion Models for Text-to-image Synthesis on the Cloud | acl-demos.1 | Poster | 2309.05534v1 |
https://aclanthology.org/2024.acl-demos.2.bib | @inproceedings{yuan-etal-2024-openvna,
title = "{O}pen{VNA}: A Framework for Analyzing the Behavior of Multimodal Language Understanding System under Noisy Scenarios",
author = "Yuan, Ziqi and
Zhang, Baozheng and
Xu, Hua and
Liang, Zhiyun and
Gao, Kai",
editor = "Cao, Yixin and
Feng, Yang and
Xiong, Deyi",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 3: System Demonstrations)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-demos.2",
pages = "9--18",
abstract = "We present OpenVNA, an open-source framework designed for analyzing the behavior of multimodal language understanding systems under noisy conditions. OpenVNA serves as an intuitive toolkit tailored for researchers, facilitating convenience batch-level robustness evaluation and on-the-fly instance-level demonstration. It primarily features a benchmark Python library for assessing global model robustness, offering high flexibility and extensibility, thereby enabling customization with user-defined noise types and models. Additionally, a GUI-based interface has been developed to intuitively analyze local model behavior. In this paper, we delineate the design principles and utilization of the created library and GUI-based web platform. Currently, OpenVNA is publicly accessible at \url{https://github.com/thuiar/OpenVNA}, with a demonstration video available at \url{https://youtu.be/0Z9cW7RGct4}.",
}
| We present OpenVNA, an open-source framework designed for analyzing the behavior of multimodal language understanding systems under noisy conditions. OpenVNA serves as an intuitive toolkit tailored for researchers, facilitating convenience batch-level robustness evaluation and on-the-fly instance-level demonstration. It primarily features a benchmark Python library for assessing global model robustness, offering high flexibility and extensibility, thereby enabling customization with user-defined noise types and models. Additionally, a GUI-based interface has been developed to intuitively analyze local model behavior. In this paper, we delineate the design principles and utilization of the created library and GUI-based web platform. Currently, OpenVNA is publicly accessible at \url{https://github.com/thuiar/OpenVNA}, with a demonstration video available at \url{https://youtu.be/0Z9cW7RGct4}. | [
"Yuan, Ziqi",
"Zhang, Baozheng",
"Xu, Hua",
"Liang, Zhiyun",
"Gao, Kai"
]
| {O}pen{VNA}: A Framework for Analyzing the Behavior of Multimodal Language Understanding System under Noisy Scenarios | acl-demos.2 | Poster | 2407.02773v1 |
https://aclanthology.org/2024.acl-demos.3.bib | @inproceedings{fei-etal-2024-xnlp,
title = "{XNLP}: An Interactive Demonstration System for Universal Structured {NLP}",
author = "Fei, Hao and
Zhang, Meishan and
Zhang, Min and
Chua, Tat-Seng",
editor = "Cao, Yixin and
Feng, Yang and
Xiong, Deyi",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 3: System Demonstrations)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-demos.3",
pages = "19--30",
abstract = "Structured Natural Language Processing (XNLP) is an important subset of NLP that entails understanding the underlying semantic or syntactic structure of texts, which serves as a foundational component for many downstream applications. Despite certain recent efforts to explore universal solutions for specific categories of XNLP tasks, a comprehensive and effective approach for unifying all XNLP tasks long remains underdeveloped. Meanwhile, while XNLP demonstration systems are vital for researchers exploring various XNLP tasks, existing platforms can be limited to, e.g., supporting few XNLP tasks, lacking interactivity and universalness. To this end, we propose an advanced XNLP demonstration system, where we leverage LLM to achieve universal XNLP, with one model for all with high generalizability. Overall, our system advances in multiple aspects, including universal XNLP modeling, high performance, interpretability, scalability, and interactivity, offering a unified platform for exploring diverse XNLP tasks in the community.",
}
| Structured Natural Language Processing (XNLP) is an important subset of NLP that entails understanding the underlying semantic or syntactic structure of texts, which serves as a foundational component for many downstream applications. Despite certain recent efforts to explore universal solutions for specific categories of XNLP tasks, a comprehensive and effective approach for unifying all XNLP tasks long remains underdeveloped. Meanwhile, while XNLP demonstration systems are vital for researchers exploring various XNLP tasks, existing platforms can be limited to, e.g., supporting few XNLP tasks, lacking interactivity and universalness. To this end, we propose an advanced XNLP demonstration system, where we leverage LLM to achieve universal XNLP, with one model for all with high generalizability. Overall, our system advances in multiple aspects, including universal XNLP modeling, high performance, interpretability, scalability, and interactivity, offering a unified platform for exploring diverse XNLP tasks in the community. | [
"Fei, Hao",
"Zhang, Meishan",
"Zhang, Min",
"Chua, Tat-Seng"
]
| {XNLP}: An Interactive Demonstration System for Universal Structured {NLP} | acl-demos.3 | Poster | 2308.01846v2 |
https://aclanthology.org/2024.acl-demos.4.bib | @inproceedings{wu-etal-2024-towards-topmost,
title = "Towards the {T}op{M}ost: A Topic Modeling System Toolkit",
author = "Wu, Xiaobao and
Pan, Fengjun and
Luu, Anh Tuan",
editor = "Cao, Yixin and
Feng, Yang and
Xiong, Deyi",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 3: System Demonstrations)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-demos.4",
pages = "31--41",
abstract = "Topic models have a rich history with various applications and have recently been reinvigorated by neural topic modeling. However, these numerous topic models adopt totally distinct datasets, implementations, and evaluations. This impedes quick utilization and fair comparisons, and thereby hinders their research progress and applications. To tackle this challenge, we in this paper propose a Topic Modeling System Toolkit (TopMost). Compared to existing toolkits, TopMost stands out by supporting more extensive features. It covers a broader spectrum of topic modeling scenarios with their complete lifecycles, including datasets, preprocessing, models, training, and evaluations. Thanks to its highly cohesive and decoupled modular design, TopMost enables rapid utilization, fair comparisons, and flexible extensions of diverse cutting-edge topic models. Our code, tutorials, and documentation are available at https://github.com/bobxwu/topmost.",
}
| Topic models have a rich history with various applications and have recently been reinvigorated by neural topic modeling. However, these numerous topic models adopt totally distinct datasets, implementations, and evaluations. This impedes quick utilization and fair comparisons, and thereby hinders their research progress and applications. To tackle this challenge, we in this paper propose a Topic Modeling System Toolkit (TopMost). Compared to existing toolkits, TopMost stands out by supporting more extensive features. It covers a broader spectrum of topic modeling scenarios with their complete lifecycles, including datasets, preprocessing, models, training, and evaluations. Thanks to its highly cohesive and decoupled modular design, TopMost enables rapid utilization, fair comparisons, and flexible extensions of diverse cutting-edge topic models. Our code, tutorials, and documentation are available at https://github.com/bobxwu/topmost. | [
"Wu, Xiaobao",
"Pan, Fengjun",
"Luu, Anh Tuan"
]
| Towards the {T}op{M}ost: A Topic Modeling System Toolkit | acl-demos.4 | Poster | 2309.06908v2 |
https://aclanthology.org/2024.acl-demos.5.bib | @inproceedings{wang-etal-2024-wordflow,
title = "Wordflow: Social Prompt Engineering for Large Language Models",
author = "Wang, Zijie and
Chakravarthy, Aishwarya and
Munechika, David and
Chau, Duen Horng",
editor = "Cao, Yixin and
Feng, Yang and
Xiong, Deyi",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 3: System Demonstrations)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-demos.5",
pages = "42--50",
abstract = "Large language models (LLMs) require well-crafted prompts for effective use. Prompt engineering, the process of designing prompts, is challenging, particularly for non-experts who are less familiar with AI technologies. While researchers have proposed techniques and tools to assist LLM users in prompt design, these works primarily target AI application developers rather than non-experts. To address this research gap, we propose social prompt engineering, a novel paradigm that leverages social computing techniques to facilitate collaborative prompt design. To investigate social prompt engineering, we introduce Wordflow, an open-source and social text editor that enables everyday users to easily create, run, share, and discover LLM prompts. Additionally, by leveraging modern web technologies, Wordflow allows users to run LLMs locally and privately in their browsers. Two usage scenarios highlight how social prompt engineering and our tool can enhance laypeople{'}s interaction with LLMs. Wordflow is publicly accessible at https://poloclub.github.io/wordflow.",
}
| Large language models (LLMs) require well-crafted prompts for effective use. Prompt engineering, the process of designing prompts, is challenging, particularly for non-experts who are less familiar with AI technologies. While researchers have proposed techniques and tools to assist LLM users in prompt design, these works primarily target AI application developers rather than non-experts. To address this research gap, we propose social prompt engineering, a novel paradigm that leverages social computing techniques to facilitate collaborative prompt design. To investigate social prompt engineering, we introduce Wordflow, an open-source and social text editor that enables everyday users to easily create, run, share, and discover LLM prompts. Additionally, by leveraging modern web technologies, Wordflow allows users to run LLMs locally and privately in their browsers. Two usage scenarios highlight how social prompt engineering and our tool can enhance laypeople{'}s interaction with LLMs. Wordflow is publicly accessible at https://poloclub.github.io/wordflow. | [
"Wang, Zijie",
"Chakravarthy, Aishwarya",
"Munechika, David",
"Chau, Duen Horng"
]
| Wordflow: Social Prompt Engineering for Large Language Models | acl-demos.5 | Poster | 2307.12980v1 |
https://aclanthology.org/2024.acl-demos.6.bib | @inproceedings{tufanov-etal-2024-lm,
title = "{LM} Transparency Tool: Interactive Tool for Analyzing Transformer Language Models",
author = "Tufanov, Igor and
Hambardzumyan, Karen and
Ferrando, Javier and
Voita, Elena",
editor = "Cao, Yixin and
Feng, Yang and
Xiong, Deyi",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 3: System Demonstrations)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-demos.6",
pages = "51--60",
abstract = "We present the LM Transparency Tool (LM-TT), an open-source interactive toolkit for analyzing the internal workings of Transformer-based language models. Differently from previously existing tools that focus on isolated parts of the decision-making process, our framework is designed to make the entire prediction process transparent, and allows tracing back model behavior from the top-layer representation to very fine-grained parts of the model. Specifically, it (i) shows the important part of the whole input-to-output information flow, (ii) allows attributing any changes done by a model block to individual attention heads and feed-forward neurons, (iii) allows interpreting the functions of those heads or neurons. A crucial part of this pipeline is showing the importance of specific model components at each step. As a result, we are able to look at the roles of model components only in cases where they are important for a prediction. Since knowing which components should be inspected is key for analyzing large models where the number of these components is extremely high, we believe our tool will greatly support the interpretability community both in research settings and in practical applications.",
}
| We present the LM Transparency Tool (LM-TT), an open-source interactive toolkit for analyzing the internal workings of Transformer-based language models. Differently from previously existing tools that focus on isolated parts of the decision-making process, our framework is designed to make the entire prediction process transparent, and allows tracing back model behavior from the top-layer representation to very fine-grained parts of the model. Specifically, it (i) shows the important part of the whole input-to-output information flow, (ii) allows attributing any changes done by a model block to individual attention heads and feed-forward neurons, (iii) allows interpreting the functions of those heads or neurons. A crucial part of this pipeline is showing the importance of specific model components at each step. As a result, we are able to look at the roles of model components only in cases where they are important for a prediction. Since knowing which components should be inspected is key for analyzing large models where the number of these components is extremely high, we believe our tool will greatly support the interpretability community both in research settings and in practical applications. | [
"Tufanov, Igor",
"Hambardzumyan, Karen",
"Ferr",
"o, Javier",
"Voita, Elena"
]
| {LM} Transparency Tool: Interactive Tool for Analyzing Transformer Language Models | acl-demos.6 | Poster | 2404.07004v1 |
https://aclanthology.org/2024.acl-demos.7.bib | @inproceedings{fei-etal-2024-empathyear,
title = "{E}mpathy{E}ar: An Open-source Avatar Multimodal Empathetic Chatbot",
author = "Fei, Hao and
Zhang, Han and
Wang, Bin and
Liao, Lizi and
Liu, Qian and
Cambria, Erik",
editor = "Cao, Yixin and
Feng, Yang and
Xiong, Deyi",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 3: System Demonstrations)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-demos.7",
pages = "61--71",
abstract = "This paper introduces EmpathyEar, a pioneering open-source, avatar-based multimodal empathetic chatbot, to fill the gap in traditional text-only empathetic response generation (ERG) systems. Leveraging the advancements of a large language model, combined with multimodal encoders and generators, EmpathyEar supports user inputs in any combination of text, sound, and vision, and produces multimodal empathetic responses, offering users, not just textual responses but also digital avatars with talking faces and synchronized speeches. A series of emotion-aware instruction-tuning is performed for comprehensive emotional understanding and generation capabilities. In this way, EmpathyEar provides users with responses that achieve a deeper emotional resonance, closely emulating human-like empathy. The system paves the way for the next emotional intelligence, for which we open-source the code for public access.",
}
| This paper introduces EmpathyEar, a pioneering open-source, avatar-based multimodal empathetic chatbot, to fill the gap in traditional text-only empathetic response generation (ERG) systems. Leveraging the advancements of a large language model, combined with multimodal encoders and generators, EmpathyEar supports user inputs in any combination of text, sound, and vision, and produces multimodal empathetic responses, offering users, not just textual responses but also digital avatars with talking faces and synchronized speeches. A series of emotion-aware instruction-tuning is performed for comprehensive emotional understanding and generation capabilities. In this way, EmpathyEar provides users with responses that achieve a deeper emotional resonance, closely emulating human-like empathy. The system paves the way for the next emotional intelligence, for which we open-source the code for public access. | [
"Fei, Hao",
"Zhang, Han",
"Wang, Bin",
"Liao, Lizi",
"Liu, Qian",
"Cambria, Erik"
]
| {E}mpathy{E}ar: An Open-source Avatar Multimodal Empathetic Chatbot | acl-demos.7 | Poster | 2406.15177v1 |
https://aclanthology.org/2024.acl-demos.8.bib | @inproceedings{iong-etal-2024-openwebagent,
title = "{O}pen{W}eb{A}gent: An Open Toolkit to Enable Web Agents on Large Language Models",
author = "Iong, Iat Long and
Liu, Xiao and
Chen, Yuxuan and
Lai, Hanyu and
Yao, Shuntian and
Shen, Pengbo and
Yu, Hao and
Dong, Yuxiao and
Tang, Jie",
editor = "Cao, Yixin and
Feng, Yang and
Xiong, Deyi",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 3: System Demonstrations)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-demos.8",
pages = "72--81",
abstract = "We introduce OpenWebAgent, an open toolkit designed to optimize web automation by integrating both large language models (LLMs) and large multimodal models (LMMs). This toolkit focuses on enhancing human-computer interactions on the web, simplifying complex tasks through an advanced HTML parser, a rapid action generation module, and an intuitive user interface. At the core of OpenWebAgent is an innovative web agent framework that uses a modular design to allow developers to seamlessly integrate a variety of models and tools to process web information and automate tasks on the web. This enables the development of powerful, task-oriented web agents, significantly enhancing user experience and operational efficiency on the web. The OpenWebAgent framework, Chrome plugin, and demo video are available at https://github.com/THUDM/OpenWebAgent/.",
}
| We introduce OpenWebAgent, an open toolkit designed to optimize web automation by integrating both large language models (LLMs) and large multimodal models (LMMs). This toolkit focuses on enhancing human-computer interactions on the web, simplifying complex tasks through an advanced HTML parser, a rapid action generation module, and an intuitive user interface. At the core of OpenWebAgent is an innovative web agent framework that uses a modular design to allow developers to seamlessly integrate a variety of models and tools to process web information and automate tasks on the web. This enables the development of powerful, task-oriented web agents, significantly enhancing user experience and operational efficiency on the web. The OpenWebAgent framework, Chrome plugin, and demo video are available at https://github.com/THUDM/OpenWebAgent/. | [
"Iong, Iat Long",
"Liu, Xiao",
"Chen, Yuxuan",
"Lai, Hanyu",
"Yao, Shuntian",
"Shen, Pengbo",
"Yu, Hao",
"Dong, Yuxiao",
"Tang, Jie"
]
| {O}pen{W}eb{A}gent: An Open Toolkit to Enable Web Agents on Large Language Models | acl-demos.8 | Poster | 2408.02248v1 |
https://aclanthology.org/2024.acl-demos.9.bib | @inproceedings{wang-etal-2024-easyedit,
title = "{E}asy{E}dit: An Easy-to-use Knowledge Editing Framework for Large Language Models",
author = "Wang, Peng and
Zhang, Ningyu and
Tian, Bozhong and
Xi, Zekun and
Yao, Yunzhi and
Xu, Ziwen and
Wang, Mengru and
Mao, Shengyu and
Wang, Xiaohan and
Cheng, Siyuan and
Liu, Kangwei and
Ni, Yuansheng and
Zheng, Guozhou and
Chen, Huajun",
editor = "Cao, Yixin and
Feng, Yang and
Xiong, Deyi",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 3: System Demonstrations)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-demos.9",
pages = "82--93",
abstract = "Large Language Models (LLMs) usually suffer from knowledge cutoff or fallacy issues, which means they are unaware of unseen events or generate text with incorrect facts owing to outdated/noisy data. To this end, many knowledge editing approaches for LLMs have emerged {--} aiming to subtly inject/edit updated knowledge or adjust undesired behavior while minimizing the impact on unrelated inputs. Nevertheless, due to significant differences among various knowledge editing methods and the variations in task setups, there is no standard implementation framework available for the community, which hinders practitioners from applying knowledge editing to applications. To address these issues, we propose EasyEdit, an easy-to-use knowledge editing framework for LLMs. It supports various cutting-edge knowledge editing approaches and can be readily applied to many well-known LLMs such as T5, GPT-J, LlaMA, etc. Empirically, we report the knowledge editing results on LlaMA-2 with EasyEdit, demonstrating that knowledge editing surpasses traditional fine-tuning in terms of reliability and generalization. We have released the source code on GitHub, along with Google Colab tutorials and comprehensive documentation for beginners to get started. Besides, we present an online system for real-time knowledge editing, and a demo video.",
}
| Large Language Models (LLMs) usually suffer from knowledge cutoff or fallacy issues, which means they are unaware of unseen events or generate text with incorrect facts owing to outdated/noisy data. To this end, many knowledge editing approaches for LLMs have emerged {--} aiming to subtly inject/edit updated knowledge or adjust undesired behavior while minimizing the impact on unrelated inputs. Nevertheless, due to significant differences among various knowledge editing methods and the variations in task setups, there is no standard implementation framework available for the community, which hinders practitioners from applying knowledge editing to applications. To address these issues, we propose EasyEdit, an easy-to-use knowledge editing framework for LLMs. It supports various cutting-edge knowledge editing approaches and can be readily applied to many well-known LLMs such as T5, GPT-J, LlaMA, etc. Empirically, we report the knowledge editing results on LlaMA-2 with EasyEdit, demonstrating that knowledge editing surpasses traditional fine-tuning in terms of reliability and generalization. We have released the source code on GitHub, along with Google Colab tutorials and comprehensive documentation for beginners to get started. Besides, we present an online system for real-time knowledge editing, and a demo video. | [
"Wang, Peng",
"Zhang, Ningyu",
"Tian, Bozhong",
"Xi, Zekun",
"Yao, Yunzhi",
"Xu, Ziwen",
"Wang, Mengru",
"Mao, Shengyu",
"Wang, Xiaohan",
"Cheng, Siyuan",
"Liu, Kangwei",
"Ni, Yuansheng",
"Zheng, Guozhou",
"Chen, Huajun"
]
| {E}asy{E}dit: An Easy-to-use Knowledge Editing Framework for Large Language Models | acl-demos.9 | Poster | 2312.05497v3 |
https://aclanthology.org/2024.acl-demos.10.bib | @inproceedings{ou-etal-2024-easyinstruct,
title = "{E}asy{I}nstruct: An Easy-to-use Instruction Processing Framework for Large Language Models",
author = "Ou, Yixin and
Zhang, Ningyu and
Gui, Honghao and
Xu, Ziwen and
Qiao, Shuofei and
Fang, Runnan and
Li, Lei and
Bi, Zhen and
Zheng, Guozhou and
Chen, Huajun",
editor = "Cao, Yixin and
Feng, Yang and
Xiong, Deyi",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 3: System Demonstrations)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-demos.10",
pages = "94--106",
abstract = "In recent years, instruction tuning has gained increasing attention and emerged as a crucial technique to enhance the capabilities of Large Language Models (LLMs). To construct high-quality instruction datasets, many instruction processing approaches have been proposed, aiming to achieve a delicate balance between data quantity and data quality. Nevertheless, due to inconsistencies that persist among various instruction processing methods, there is no standard open-source instruction processing implementation framework available for the community, which hinders practitioners from further developing and advancing. To facilitate instruction processing research and development, we present EasyInstruct, an easy-to-use instruction processing framework for LLMs, which modularizes instruction generation, selection, and prompting, while also considering their combination and interaction. EasyInstruct is publicly released and actively maintained at Github, along with an online demo app and a demo video for quick-start, calling for broader research centered on instruction data and synthetic data.",
}
| In recent years, instruction tuning has gained increasing attention and emerged as a crucial technique to enhance the capabilities of Large Language Models (LLMs). To construct high-quality instruction datasets, many instruction processing approaches have been proposed, aiming to achieve a delicate balance between data quantity and data quality. Nevertheless, due to inconsistencies that persist among various instruction processing methods, there is no standard open-source instruction processing implementation framework available for the community, which hinders practitioners from further developing and advancing. To facilitate instruction processing research and development, we present EasyInstruct, an easy-to-use instruction processing framework for LLMs, which modularizes instruction generation, selection, and prompting, while also considering their combination and interaction. EasyInstruct is publicly released and actively maintained at Github, along with an online demo app and a demo video for quick-start, calling for broader research centered on instruction data and synthetic data. | [
"Ou, Yixin",
"Zhang, Ningyu",
"Gui, Honghao",
"Xu, Ziwen",
"Qiao, Shuofei",
"Fang, Runnan",
"Li, Lei",
"Bi, Zhen",
"Zheng, Guozhou",
"Chen, Huajun"
]
| {E}asy{I}nstruct: An Easy-to-use Instruction Processing Framework for Large Language Models | acl-demos.10 | Poster | 2311.18215v1 |
https://aclanthology.org/2024.acl-demos.11.bib | @inproceedings{cho-etal-2024-boteval,
title = "{B}ot{E}val: Facilitating Interactive Human Evaluation",
author = "Cho, Hyundong and
Gowda, Thamme and
Huang, Yuyang and
Lu, Zixun and
Tong, Tianli and
May, Jonathan",
editor = "Cao, Yixin and
Feng, Yang and
Xiong, Deyi",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 3: System Demonstrations)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-demos.11",
pages = "107--116",
abstract = "Following the rapid progress in natural language processing (NLP) models, language models are applied to increasingly more complex interactive tasks such as negotiations and conversation moderations. Having human evaluators directly interact with these NLP models is essential for adequately evaluating the performance on such interactive tasks. We develop BotEval, an easily customizable, open-source, evaluation toolkit that focuses on enabling human-bot interactions as part of the evaluation process, as opposed to human evaluators making judgements for a static input. BotEval balances flexibility for customization and user-friendliness by providing templates for common use cases that span various degrees of complexity and built-in compatibility with popular crowdsourcing platforms.We showcase the numerous useful features of BotEval through a study that evaluates the performance of various chatbots on their effectiveness for conversational moderation and discuss how BotEval differs from other annotation tools.",
}
| Following the rapid progress in natural language processing (NLP) models, language models are applied to increasingly more complex interactive tasks such as negotiations and conversation moderations. Having human evaluators directly interact with these NLP models is essential for adequately evaluating the performance on such interactive tasks. We develop BotEval, an easily customizable, open-source, evaluation toolkit that focuses on enabling human-bot interactions as part of the evaluation process, as opposed to human evaluators making judgements for a static input. BotEval balances flexibility for customization and user-friendliness by providing templates for common use cases that span various degrees of complexity and built-in compatibility with popular crowdsourcing platforms.We showcase the numerous useful features of BotEval through a study that evaluates the performance of various chatbots on their effectiveness for conversational moderation and discuss how BotEval differs from other annotation tools. | [
"Cho, Hyundong",
"Gowda, Thamme",
"Huang, Yuyang",
"Lu, Zixun",
"Tong, Tianli",
"May, Jonathan"
]
| {B}ot{E}val: Facilitating Interactive Human Evaluation | acl-demos.11 | Poster | 2109.14700v2 |
https://aclanthology.org/2024.acl-demos.12.bib | @inproceedings{takeshita-etal-2024-gengo,
title = "{G}en{GO}: {ACL} Paper Explorer with Semantic Features",
author = "Takeshita, Sotaro and
Ponzetto, Simone and
Eckert, Kai",
editor = "Cao, Yixin and
Feng, Yang and
Xiong, Deyi",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 3: System Demonstrations)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-demos.12",
pages = "117--126",
abstract = "We present GenGO, a system for exploring papers published in ACL conferences. Paper data stored in our database is enriched with multi-aspect summaries, extracted named entities, a field of study label, and text embeddings by our data processing pipeline. These metadata are used in our web-based user interface to enable researchers to quickly find papers relevant to their interests, and grasp an overview of papers without reading full-text of papers. To make GenGO to be available online as long as possible, we design GenGO to be simple and efficient to reduce maintenance and financial costs. In addition, the modularity of our data processing pipeline lets developers easily extend it to add new features. We make our code available to foster open development and transparency: https://gengo.sotaro.io.",
}
| We present GenGO, a system for exploring papers published in ACL conferences. Paper data stored in our database is enriched with multi-aspect summaries, extracted named entities, a field of study label, and text embeddings by our data processing pipeline. These metadata are used in our web-based user interface to enable researchers to quickly find papers relevant to their interests, and grasp an overview of papers without reading full-text of papers. To make GenGO to be available online as long as possible, we design GenGO to be simple and efficient to reduce maintenance and financial costs. In addition, the modularity of our data processing pipeline lets developers easily extend it to add new features. We make our code available to foster open development and transparency: https://gengo.sotaro.io. | [
"Takeshita, Sotaro",
"Ponzetto, Simone",
"Eckert, Kai"
]
| {G}en{GO}: {ACL} Paper Explorer with Semantic Features | acl-demos.12 | Poster | 2405.16011v1 |
https://aclanthology.org/2024.acl-demos.13.bib | @inproceedings{schopf-matthes-2024-nlp,
title = "{NLP}-{KG}: A System for Exploratory Search of Scientific Literature in Natural Language Processing",
author = "Schopf, Tim and
Matthes, Florian",
editor = "Cao, Yixin and
Feng, Yang and
Xiong, Deyi",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 3: System Demonstrations)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-demos.13",
pages = "127--135",
abstract = "Scientific literature searches are often exploratory, whereby users are not yet familiar with a particular field or concept but are interested in learning more about it. However, existing systems for scientific literature search are typically tailored to keyword-based lookup searches, limiting the possibilities for exploration. We propose NLP-KG, a feature-rich system designed to support the exploration of research literature in unfamiliar natural language processing (NLP) fields. In addition to a semantic search, NLP-KG allows users to easily find survey papers that provide a quick introduction to a field of interest. Further, a Fields of Study hierarchy graph enables users to familiarize themselves with a field and its related areas. Finally, a chat interface allows users to ask questions about unfamiliar concepts or specific articles in NLP and obtain answers grounded in knowledge retrieved from scientific publications. Our system provides users with comprehensive exploration possibilities, supporting them in investigating the relationships between different fields, understanding unfamiliar concepts in NLP, and finding relevant research literature. Demo, video, and code are available at: https://github.com/NLP-Knowledge-Graph/NLP-KG-WebApp.",
}
| Scientific literature searches are often exploratory, whereby users are not yet familiar with a particular field or concept but are interested in learning more about it. However, existing systems for scientific literature search are typically tailored to keyword-based lookup searches, limiting the possibilities for exploration. We propose NLP-KG, a feature-rich system designed to support the exploration of research literature in unfamiliar natural language processing (NLP) fields. In addition to a semantic search, NLP-KG allows users to easily find survey papers that provide a quick introduction to a field of interest. Further, a Fields of Study hierarchy graph enables users to familiarize themselves with a field and its related areas. Finally, a chat interface allows users to ask questions about unfamiliar concepts or specific articles in NLP and obtain answers grounded in knowledge retrieved from scientific publications. Our system provides users with comprehensive exploration possibilities, supporting them in investigating the relationships between different fields, understanding unfamiliar concepts in NLP, and finding relevant research literature. Demo, video, and code are available at: https://github.com/NLP-Knowledge-Graph/NLP-KG-WebApp. | [
"Schopf, Tim",
"Matthes, Florian"
]
| {NLP}-{KG}: A System for Exploratory Search of Scientific Literature in Natural Language Processing | acl-demos.13 | Poster | 2406.15294v2 |
https://aclanthology.org/2024.acl-demos.14.bib | @inproceedings{yu-etal-2024-localrqa,
title = "{L}ocal{RQA}: From Generating Data to Locally Training, Testing, and Deploying Retrieval-Augmented {QA} Systems",
author = "Yu, Xiao and
Lu, Yunan and
Yu, Zhou",
editor = "Cao, Yixin and
Feng, Yang and
Xiong, Deyi",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 3: System Demonstrations)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-demos.14",
pages = "136--151",
abstract = "Retrieval-augmented question-answering systems combine retrieval techniques with large language models to provide answers that are more accurate and informative. Many existing toolkits allow users to quickly build such systems using off-the-shelf models, but they fall short in supporting researchers and developers to customize the *model training, testing, and deployment process*. We propose LocalRQA, an open-source toolkit that features a wide selection of model training algorithms, evaluation methods, and deployment tools curated from the latest research. As a showcase, we build QA systems using online documentation obtained from Databricks and Faire{'}s websites. We find 7B-models trained and deployed using LocalRQA reach a similar performance compared to using OpenAI{'}s text-ada-002 and GPT-4-turbo.",
}
| Retrieval-augmented question-answering systems combine retrieval techniques with large language models to provide answers that are more accurate and informative. Many existing toolkits allow users to quickly build such systems using off-the-shelf models, but they fall short in supporting researchers and developers to customize the *model training, testing, and deployment process*. We propose LocalRQA, an open-source toolkit that features a wide selection of model training algorithms, evaluation methods, and deployment tools curated from the latest research. As a showcase, we build QA systems using online documentation obtained from Databricks and Faire{'}s websites. We find 7B-models trained and deployed using LocalRQA reach a similar performance compared to using OpenAI{'}s text-ada-002 and GPT-4-turbo. | [
"Yu, Xiao",
"Lu, Yunan",
"Yu, Zhou"
]
| {L}ocal{RQA}: From Generating Data to Locally Training, Testing, and Deploying Retrieval-Augmented {QA} Systems | acl-demos.14 | Poster | 2403.00982v1 |
https://aclanthology.org/2024.acl-demos.15.bib | @inproceedings{tahir-etal-2024-jora,
title = "{JORA}: {JAX} Tensor-Parallel {L}o{RA} Library for Retrieval Augmented Fine-Tuning",
author = "Tahir, Anique and
Cheng, Lu and
Liu, Huan",
editor = "Cao, Yixin and
Feng, Yang and
Xiong, Deyi",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 3: System Demonstrations)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-demos.15",
pages = "152--159",
abstract = "The scaling of Large Language Models (LLMs) for retrieval-based tasks, particularly in Retrieval Augmented Generation (RAG), faces significant memory constraints, especially when fine-tuning extensive prompt sequences. Current open-source libraries support full-model inference and fine-tuning across multiple GPUs but fall short of accommodating the efficient parameter distribution required for retrieved context. Addressing this gap, we introduce a novel framework for PEFT-compatible fine-tuning of GPT models, leveraging distributed training. Our framework uniquely utilizes JAX{'}s just-in-time (JIT) compilation and tensor-sharding for efficient resource management, thereby enabling accelerated fine-tuning with reduced memory requirements. This advancement significantly improves the scalability and feasibility of fine-tuning LLMs for complex RAG applications, even on systems with limited GPU resources. Our experiments show more than 12x improvement in runtime compared to Hugging Face/DeepSpeed implementation with four GPUs while consuming less than half the VRAM per GPU.",
}
| The scaling of Large Language Models (LLMs) for retrieval-based tasks, particularly in Retrieval Augmented Generation (RAG), faces significant memory constraints, especially when fine-tuning extensive prompt sequences. Current open-source libraries support full-model inference and fine-tuning across multiple GPUs but fall short of accommodating the efficient parameter distribution required for retrieved context. Addressing this gap, we introduce a novel framework for PEFT-compatible fine-tuning of GPT models, leveraging distributed training. Our framework uniquely utilizes JAX{'}s just-in-time (JIT) compilation and tensor-sharding for efficient resource management, thereby enabling accelerated fine-tuning with reduced memory requirements. This advancement significantly improves the scalability and feasibility of fine-tuning LLMs for complex RAG applications, even on systems with limited GPU resources. Our experiments show more than 12x improvement in runtime compared to Hugging Face/DeepSpeed implementation with four GPUs while consuming less than half the VRAM per GPU. | [
"Tahir, Anique",
"Cheng, Lu",
"Liu, Huan"
]
| {JORA}: {JAX} Tensor-Parallel {L}o{RA} Library for Retrieval Augmented Fine-Tuning | acl-demos.15 | Poster | 2406.16989v2 |
https://aclanthology.org/2024.acl-demos.16.bib | @inproceedings{zhao-etal-2024-lingualinked,
title = "{L}ingua{L}inked: Distributed Large Language Model Inference on Mobile Devices",
author = "Zhao, Junchen and
Song, Yurun and
[email protected], [email protected] and
Harris, Ian and
Abdu Jyothi, Sangeetha",
editor = "Cao, Yixin and
Feng, Yang and
Xiong, Deyi",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 3: System Demonstrations)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-demos.16",
pages = "160--171",
abstract = "Deploying Large Language Models (LLMs) locally on mobile devices presents a significant challenge due to their extensive memory requirements. In this paper, we introduce LinguaLinked, a system for decentralized, distributed LLM inference on mobile devices. LinguaLinked enables collaborative execution of the inference task across multiple trusted devices and ensures data privacy by processing information locally. LinguaLinked uses three key strategies. First, an optimized model assignment technique segments LLMs and uses linear optimization to align segments with each device�s capabilities. Second, an optimized data transmission mechanism ensures efficient and structured data flow between model segments while also maintaining the integrity of the original model structure. Finally, LinguaLinked incorporates a runtime load balancer that actively monitors and redistributes tasks among mobile devices to prevent bottlenecks, enhancing the system�s overall efficiency and responsiveness. We demonstrate that LinguaLinked facilitates efficient LLM inference while maintaining consistent throughput and minimal latency through extensive testing across various mobile devices, from high-end to low-end Android devices.",
}
| Deploying Large Language Models (LLMs) locally on mobile devices presents a significant challenge due to their extensive memory requirements. In this paper, we introduce LinguaLinked, a system for decentralized, distributed LLM inference on mobile devices. LinguaLinked enables collaborative execution of the inference task across multiple trusted devices and ensures data privacy by processing information locally. LinguaLinked uses three key strategies. First, an optimized model assignment technique segments LLMs and uses linear optimization to align segments with each device�s capabilities. Second, an optimized data transmission mechanism ensures efficient and structured data flow between model segments while also maintaining the integrity of the original model structure. Finally, LinguaLinked incorporates a runtime load balancer that actively monitors and redistributes tasks among mobile devices to prevent bottlenecks, enhancing the system�s overall efficiency and responsiveness. We demonstrate that LinguaLinked facilitates efficient LLM inference while maintaining consistent throughput and minimal latency through extensive testing across various mobile devices, from high-end to low-end Android devices. | [
"Zhao, Junchen",
"Song, Yurun",
"[email protected], [email protected]",
"Harris, Ian",
"Abdu Jyothi, Sangeetha"
]
| {L}ingua{L}inked: Distributed Large Language Model Inference on Mobile Devices | acl-demos.16 | Poster | 2312.00388v1 |
https://aclanthology.org/2024.acl-demos.17.bib | @inproceedings{spiegel-macko-2024-imgtb,
title = "{IMGTB}: A Framework for Machine-Generated Text Detection Benchmarking",
author = "Spiegel, Michal and
Macko, Dominik",
editor = "Cao, Yixin and
Feng, Yang and
Xiong, Deyi",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 3: System Demonstrations)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-demos.17",
pages = "172--179",
abstract = "In the era of large language models generating high quality texts, it is a necessity to develop methods for detection of machine-generated text to avoid their harmful use or simply for annotation purposes. It is, however, also important to properly evaluate and compare such developed methods. Recently, a few benchmarks have been proposed for this purpose; however, integration of newest detection methods is rather challenging, since new methods appear each month and provide slightly different evaluation pipelines.In this paper, we present the IMGTB framework, which simplifies the benchmarking of machine-generated text detection methods by easy integration of custom (new) methods and evaluation datasets. In comparison to existing frameworks, it enables to objectively compare statistical metric-based zero-shot detectors with classification-based detectors and with differently fine-tuned detectors. Its configurability and flexibility makes research and development of new detection methods easier, especially their comparison to the existing state-of-the-art detectors. The default set of analyses, metrics and visualizations offered by the tool follows the established practices of machine-generated text detection benchmarking found in state-of-the-art literature.",
}
| In the era of large language models generating high quality texts, it is a necessity to develop methods for detection of machine-generated text to avoid their harmful use or simply for annotation purposes. It is, however, also important to properly evaluate and compare such developed methods. Recently, a few benchmarks have been proposed for this purpose; however, integration of newest detection methods is rather challenging, since new methods appear each month and provide slightly different evaluation pipelines.In this paper, we present the IMGTB framework, which simplifies the benchmarking of machine-generated text detection methods by easy integration of custom (new) methods and evaluation datasets. In comparison to existing frameworks, it enables to objectively compare statistical metric-based zero-shot detectors with classification-based detectors and with differently fine-tuned detectors. Its configurability and flexibility makes research and development of new detection methods easier, especially their comparison to the existing state-of-the-art detectors. The default set of analyses, metrics and visualizations offered by the tool follows the established practices of machine-generated text detection benchmarking found in state-of-the-art literature. | [
"Spiegel, Michal",
"Macko, Dominik"
]
| {IMGTB}: A Framework for Machine-Generated Text Detection Benchmarking | acl-demos.17 | Poster | 2012.05030v2 |
https://aclanthology.org/2024.acl-demos.18.bib | @inproceedings{bobrov-etal-2024-drugwatch,
title = "{D}rug{W}atch: A Comprehensive Multi-Source Data Visualisation Platform for Drug Safety Information",
author = "Bobrov, Artem and
Saltenis, Domantas and
Sun, Zhaoyue and
Pergola, Gabriele and
He, Yulan",
editor = "Cao, Yixin and
Feng, Yang and
Xiong, Deyi",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 3: System Demonstrations)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-demos.18",
pages = "180--189",
abstract = "Drug safety research is crucial for maintaining public health, often requiring comprehensive data support. However, the resources currently available to the public are limited and fail to provide a comprehensive understanding of the relationship between drugs and their side effects. This paper introduces {``}DrugWatch{''}, an easy-to-use and interactive multi-source information visualisation platform for drug safety study. It allows users to understand common side effects of drugs and their statistical information, flexibly retrieve relevant medical reports, or annotate their own medical texts with our automated annotation tool. Supported by NLP technology and enriched with interactive visual components, we are committed to providing researchers and practitioners with a one-stop information analysis, retrieval, and annotation service. The demonstration video is available at https://www.youtube.com/watch?v=RTqDgxzETjw. We also deployed an online demonstration system at https://drugwatch.net/.",
}
| Drug safety research is crucial for maintaining public health, often requiring comprehensive data support. However, the resources currently available to the public are limited and fail to provide a comprehensive understanding of the relationship between drugs and their side effects. This paper introduces {``}DrugWatch{''}, an easy-to-use and interactive multi-source information visualisation platform for drug safety study. It allows users to understand common side effects of drugs and their statistical information, flexibly retrieve relevant medical reports, or annotate their own medical texts with our automated annotation tool. Supported by NLP technology and enriched with interactive visual components, we are committed to providing researchers and practitioners with a one-stop information analysis, retrieval, and annotation service. The demonstration video is available at https://www.youtube.com/watch?v=RTqDgxzETjw. We also deployed an online demonstration system at https://drugwatch.net/. | [
"Bobrov, Artem",
"Saltenis, Domantas",
"Sun, Zhaoyue",
"Pergola, Gabriele",
"He, Yulan"
]
| {D}rug{W}atch: A Comprehensive Multi-Source Data Visualisation Platform for Drug Safety Information | acl-demos.18 | Poster | 2407.01585v1 |
https://aclanthology.org/2024.acl-demos.19.bib | @inproceedings{liu-etal-2024-openeval,
title = "{O}pen{E}val: Benchmarking {C}hinese {LLM}s across Capability, Alignment and Safety",
author = "Liu, Chuang and
Yu, Linhao and
Li, Jiaxuan and
Jin, Renren and
Huang, Yufei and
Shi, Ling and
Zhang, Junhui and
Ji, Xinmeng and
Cui, Tingting and
Liutao, Liutao and
Song, Jinwang and
Zan, Hongying and
Li, Sun and
Xiong, Deyi",
editor = "Cao, Yixin and
Feng, Yang and
Xiong, Deyi",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 3: System Demonstrations)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-demos.19",
pages = "190--210",
abstract = "The rapid development of Chinese large language models (LLMs) poses big challenges for efficient LLM evaluation. While current initiatives have introduced new benchmarks or evaluation platforms for assessing Chinese LLMs, many of these focus primarily on capabilities, usually overlooking potential alignment and safety issues. To address this gap, we introduce OpenEval, an evaluation testbed that benchmarks Chinese LLMs across capability, alignment and safety. For capability assessment, we include 12 benchmark datasets to evaluate Chinese LLMs from 4 sub-dimensions: NLP tasks, disciplinary knowledge, commonsense reasoning and mathematical reasoning. For alignment assessment, OpenEval contains 7 datasets that examines the bias, offensiveness and illegalness in the outputs yielded by Chinese LLMs. To evaluate safety, especially anticipated risks (e.g., power-seeking, self-awareness) of advanced LLMs, we include 6 datasets. In addition to these benchmarks, we have implemented a phased public evaluation and benchmark update strategy to ensure that OpenEval is in line with the development of Chinese LLMs or even able to provide cutting-edge benchmark datasets to guide the development of Chinese LLMs. In our first public evaluation, we have tested a range of Chinese LLMs, spanning from 7B to 72B parameters, including both open-source and proprietary models. Evaluation results indicate that while Chinese LLMs have shown impressive performance in certain tasks, more attention should be directed towards broader aspects such as commonsense reasoning, alignment, and safety.",
}
| The rapid development of Chinese large language models (LLMs) poses big challenges for efficient LLM evaluation. While current initiatives have introduced new benchmarks or evaluation platforms for assessing Chinese LLMs, many of these focus primarily on capabilities, usually overlooking potential alignment and safety issues. To address this gap, we introduce OpenEval, an evaluation testbed that benchmarks Chinese LLMs across capability, alignment and safety. For capability assessment, we include 12 benchmark datasets to evaluate Chinese LLMs from 4 sub-dimensions: NLP tasks, disciplinary knowledge, commonsense reasoning and mathematical reasoning. For alignment assessment, OpenEval contains 7 datasets that examines the bias, offensiveness and illegalness in the outputs yielded by Chinese LLMs. To evaluate safety, especially anticipated risks (e.g., power-seeking, self-awareness) of advanced LLMs, we include 6 datasets. In addition to these benchmarks, we have implemented a phased public evaluation and benchmark update strategy to ensure that OpenEval is in line with the development of Chinese LLMs or even able to provide cutting-edge benchmark datasets to guide the development of Chinese LLMs. In our first public evaluation, we have tested a range of Chinese LLMs, spanning from 7B to 72B parameters, including both open-source and proprietary models. Evaluation results indicate that while Chinese LLMs have shown impressive performance in certain tasks, more attention should be directed towards broader aspects such as commonsense reasoning, alignment, and safety. | [
"Liu, Chuang",
"Yu, Linhao",
"Li, Jiaxuan",
"Jin, Renren",
"Huang, Yufei",
"Shi, Ling",
"Zhang, Junhui",
"Ji, Xinmeng",
"Cui, Tingting",
"Liutao, Liutao",
"Song, Jinwang",
"Zan, Hongying",
"Li, Sun",
"Xiong, Deyi"
]
| {O}pen{E}val: Benchmarking {C}hinese {LLM}s across Capability, Alignment and Safety | acl-demos.19 | Poster | 2403.12316v1 |
https://aclanthology.org/2024.acl-demos.20.bib | @inproceedings{xue-etal-2024-autore,
title = "{A}uto{RE}: Document-Level Relation Extraction with Large Language Models",
author = "Xue, Lilong and
Zhang, Dan and
Dong, Yuxiao and
Tang, Jie",
editor = "Cao, Yixin and
Feng, Yang and
Xiong, Deyi",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 3: System Demonstrations)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-demos.20",
pages = "211--220",
abstract = "Large Language Models (LLMs) have demonstrated exceptional abilities in comprehending and generating text, motivating numerous researchers to utilize them for Information Extraction (IE) purposes, including Relation Extraction (RE). Nonetheless, most existing methods are predominantly designed for Sentence-level Relation Extraction (SentRE) tasks, which typically encompass a restricted set of relations and triplet facts within a single sentence. Furthermore, certain approaches resort to treating relations as candidate choices integrated into prompt templates, leading to inefficient processing and suboptimal performance when tackling Document-Level Relation Extraction (DocRE) tasks, which entail handling multiple relations and triplet facts distributed across a given document, posing distinct challenges. To overcome these limitations, we introduce AutoRE, an end-to-end DocRE model that adopts a novel RE extraction paradigm named RHF (Relation-Head-Facts). Unlike existing approaches, AutoRE does not rely on the assumption of known relation options, making it more reflective of real-world scenarios. Additionally, we have developed an easily extensible RE framework using a Parameters Efficient Fine Tuning (PEFT) algorithm (QLoRA). Our experiments on the RE-DocRED dataset showcase AutoRE{'}s best performance, achieving state-of-the-art results, surpassing TAG by 10.03{\%} and 9.03{\%} respectively on the dev and test set. The code is available and the demonstration video is provided.",
}
| Large Language Models (LLMs) have demonstrated exceptional abilities in comprehending and generating text, motivating numerous researchers to utilize them for Information Extraction (IE) purposes, including Relation Extraction (RE). Nonetheless, most existing methods are predominantly designed for Sentence-level Relation Extraction (SentRE) tasks, which typically encompass a restricted set of relations and triplet facts within a single sentence. Furthermore, certain approaches resort to treating relations as candidate choices integrated into prompt templates, leading to inefficient processing and suboptimal performance when tackling Document-Level Relation Extraction (DocRE) tasks, which entail handling multiple relations and triplet facts distributed across a given document, posing distinct challenges. To overcome these limitations, we introduce AutoRE, an end-to-end DocRE model that adopts a novel RE extraction paradigm named RHF (Relation-Head-Facts). Unlike existing approaches, AutoRE does not rely on the assumption of known relation options, making it more reflective of real-world scenarios. Additionally, we have developed an easily extensible RE framework using a Parameters Efficient Fine Tuning (PEFT) algorithm (QLoRA). Our experiments on the RE-DocRED dataset showcase AutoRE{'}s best performance, achieving state-of-the-art results, surpassing TAG by 10.03{\%} and 9.03{\%} respectively on the dev and test set. The code is available and the demonstration video is provided. | [
"Xue, Lilong",
"Zhang, Dan",
"Dong, Yuxiao",
"Tang, Jie"
]
| {A}uto{RE}: Document-Level Relation Extraction with Large Language Models | acl-demos.20 | Poster | 2404.09593v1 |
https://aclanthology.org/2024.acl-demos.21.bib | @inproceedings{arora-dell-2024-linktransformer,
title = "{L}ink{T}ransformer: A Unified Package for Record Linkage with Transformer Language Models",
author = "Arora, Abhishek and
Dell, Melissa",
editor = "Cao, Yixin and
Feng, Yang and
Xiong, Deyi",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 3: System Demonstrations)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-demos.21",
pages = "221--231",
abstract = "Many computational analyses require linking information across noisy text datasets. While large language models (LLMs) offer significant promise, approximate string matching packages in popular statistical softwares such as R and Stata remain predominant in academic applications. These packages have simple interfaces and can be easily extended to a diversity of languages and settings, and for academic applications, ease-of-use and extensibility are essential. In contrast, packages for record linkage with LLMs require significant familiarity with deep learning frameworks and often focus on specialized applications of commercial value in English. The open-source package LinkTransformer aims to bridge this gap by providing an end-to-end software for performing record linkage and other data cleaning tasks with transformer LLMs, treating linkage as a text retrieval problem. At its core is an off-the-shelf toolkit for applying transformer models to record linkage. LinkTransformer contains a rich repository of pre-trained models for multiple languages and supports easy integration of any transformer language model from Hugging Face or OpenAI, providing the extensibility required for many scholarly applications. Its APIs also perform common data processing tasks, e.g., aggregation, noisy de-duplication, and translation-free cross-lingual linkage. LinkTransformer contains comprehensive tools for efficient model tuning, allowing for highly customized applications, and users can easily contribute their custom-trained models to its model hub to ensure reproducibility. Using a novel benchmark dataset geared towards academic applications, we show that LinkTransformer - with both custom models and Hugging Face or OpenAI models off-the-shelf - outperforms string matching by a wide margin. By combining transformer LMs with intuitive APIs, LinkTransformer aims to democratize these performance gains for those who lack familiarity with deep learning frameworks.",
}
| Many computational analyses require linking information across noisy text datasets. While large language models (LLMs) offer significant promise, approximate string matching packages in popular statistical softwares such as R and Stata remain predominant in academic applications. These packages have simple interfaces and can be easily extended to a diversity of languages and settings, and for academic applications, ease-of-use and extensibility are essential. In contrast, packages for record linkage with LLMs require significant familiarity with deep learning frameworks and often focus on specialized applications of commercial value in English. The open-source package LinkTransformer aims to bridge this gap by providing an end-to-end software for performing record linkage and other data cleaning tasks with transformer LLMs, treating linkage as a text retrieval problem. At its core is an off-the-shelf toolkit for applying transformer models to record linkage. LinkTransformer contains a rich repository of pre-trained models for multiple languages and supports easy integration of any transformer language model from Hugging Face or OpenAI, providing the extensibility required for many scholarly applications. Its APIs also perform common data processing tasks, e.g., aggregation, noisy de-duplication, and translation-free cross-lingual linkage. LinkTransformer contains comprehensive tools for efficient model tuning, allowing for highly customized applications, and users can easily contribute their custom-trained models to its model hub to ensure reproducibility. Using a novel benchmark dataset geared towards academic applications, we show that LinkTransformer - with both custom models and Hugging Face or OpenAI models off-the-shelf - outperforms string matching by a wide margin. By combining transformer LMs with intuitive APIs, LinkTransformer aims to democratize these performance gains for those who lack familiarity with deep learning frameworks. | [
"Arora, Abhishek",
"Dell, Melissa"
]
| {L}ink{T}ransformer: A Unified Package for Record Linkage with Transformer Language Models | acl-demos.21 | Poster | 2309.00789v2 |
https://aclanthology.org/2024.acl-demos.22.bib | @inproceedings{mathur-etal-2024-docpilot,
title = "{D}oc{P}ilot: Copilot for Automating {PDF} Edit Workflows in Documents",
author = "Mathur, Puneet and
Siu, Alexa and
Manjunatha, Varun and
Sun, Tong",
editor = "Cao, Yixin and
Feng, Yang and
Xiong, Deyi",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 3: System Demonstrations)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-demos.22",
pages = "232--246",
abstract = "Digital documents, such as PDFs, are vital in business workflows, enabling communication, documentation, and collaboration. Handling PDFs can involve navigating complex workflows and numerous tools (e.g., comprehension, annotation, editing), which can be tedious and time-consuming for users. We introduce DocPilot, an AI-assisted document workflow Copilot system capable of understanding user intent and executing tasks accordingly to help users streamline their workflows. DocPilot undertakes intelligent orchestration of various tools through LLM prompting in four steps: (1) Task plan generation, (2) Task plan verification and self-correction, (3) Multi-turn User Feedback, and (4) Task Plan Execution via Code Generation and Error log-based Code Self-Revision. The primary goal of this system is to free the user from the intricacies of document editing, enabling them to focus on the creative aspects and enrich their document management experience.",
}
| Digital documents, such as PDFs, are vital in business workflows, enabling communication, documentation, and collaboration. Handling PDFs can involve navigating complex workflows and numerous tools (e.g., comprehension, annotation, editing), which can be tedious and time-consuming for users. We introduce DocPilot, an AI-assisted document workflow Copilot system capable of understanding user intent and executing tasks accordingly to help users streamline their workflows. DocPilot undertakes intelligent orchestration of various tools through LLM prompting in four steps: (1) Task plan generation, (2) Task plan verification and self-correction, (3) Multi-turn User Feedback, and (4) Task Plan Execution via Code Generation and Error log-based Code Self-Revision. The primary goal of this system is to free the user from the intricacies of document editing, enabling them to focus on the creative aspects and enrich their document management experience. | [
"Mathur, Puneet",
"Siu, Alexa",
"Manjunatha, Varun",
"Sun, Tong"
]
| {D}oc{P}ilot: Copilot for Automating {PDF} Edit Workflows in Documents | acl-demos.22 | Poster | 2405.05438v1 |
https://aclanthology.org/2024.acl-demos.23.bib | @inproceedings{he-etal-2024-ultraeval,
title = "{U}ltra{E}val: A Lightweight Platform for Flexible and Comprehensive Evaluation for {LLM}s",
author = "He, Chaoqun and
Luo, Renjie and
Hu, Shengding and
Zhao, Ranchi and
Zhou, Jie and
Wu, Hanghao and
Zhang, Jiajie and
Han, Xu and
Liu, Zhiyuan and
Sun, Maosong",
editor = "Cao, Yixin and
Feng, Yang and
Xiong, Deyi",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 3: System Demonstrations)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-demos.23",
pages = "247--257",
abstract = "Evaluation is pivotal for honing Large Language Models (LLMs), pinpointing their capabilities and guiding enhancements. The rapid development of LLMs calls for a lightweight and easy-to-use framework for swift evaluation deployment. However, due to the various implementation details to consider, developing a comprehensive evaluation platform is never easy. Existing platforms are often complex and poorly modularized, hindering seamless incorporation into researcher{'}s workflows. This paper introduces UltraEval, a user-friendly evaluation framework characterized by lightweight, comprehensiveness, modularity, and efficiency. We identify and reimplement three core components of model evaluation (models, data, and metrics). The resulting composability allows for the free combination of different models, tasks, prompts, and metrics within a unified evaluation workflow. Additionally, UltraEval supports diverse models owing to a unified HTTP service and provides sufficient inference acceleration.",
}
| Evaluation is pivotal for honing Large Language Models (LLMs), pinpointing their capabilities and guiding enhancements. The rapid development of LLMs calls for a lightweight and easy-to-use framework for swift evaluation deployment. However, due to the various implementation details to consider, developing a comprehensive evaluation platform is never easy. Existing platforms are often complex and poorly modularized, hindering seamless incorporation into researcher{'}s workflows. This paper introduces UltraEval, a user-friendly evaluation framework characterized by lightweight, comprehensiveness, modularity, and efficiency. We identify and reimplement three core components of model evaluation (models, data, and metrics). The resulting composability allows for the free combination of different models, tasks, prompts, and metrics within a unified evaluation workflow. Additionally, UltraEval supports diverse models owing to a unified HTTP service and provides sufficient inference acceleration. | [
"He, Chaoqun",
"Luo, Renjie",
"Hu, Shengding",
"Zhao, Ranchi",
"Zhou, Jie",
"Wu, Hanghao",
"Zhang, Jiajie",
"Han, Xu",
"Liu, Zhiyuan",
"Sun, Maosong"
]
| {U}ltra{E}val: A Lightweight Platform for Flexible and Comprehensive Evaluation for {LLM}s | acl-demos.23 | Poster | 2404.07584v3 |
https://aclanthology.org/2024.acl-demos.24.bib | @inproceedings{hulden-etal-2024-pyfoma,
title = "{P}y{F}oma: a Python finite-state compiler module",
author = "Hulden, Mans and
Ginn, Michael and
Silfverberg, Miikka and
Hammond, Michael",
editor = "Cao, Yixin and
Feng, Yang and
Xiong, Deyi",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 3: System Demonstrations)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-demos.24",
pages = "258--265",
abstract = "We describe PyFoma, an open-source Python module for constructing weighted and unweighted finite-state transducers and automata from regular expressions, string rewriting rules, right-linear grammars, or low-level state/transition manipulation. A large variety of standard algorithms for working with finite-state machines is included, with a particular focus on the needs of linguistic and NLP applications. The data structures and code in the module are designed for legibility to allow for potential use in teaching the theory and algorithms associated with finite-state machines.",
}
| We describe PyFoma, an open-source Python module for constructing weighted and unweighted finite-state transducers and automata from regular expressions, string rewriting rules, right-linear grammars, or low-level state/transition manipulation. A large variety of standard algorithms for working with finite-state machines is included, with a particular focus on the needs of linguistic and NLP applications. The data structures and code in the module are designed for legibility to allow for potential use in teaching the theory and algorithms associated with finite-state machines. | [
"Hulden, Mans",
"Ginn, Michael",
"Silfverberg, Miikka",
"Hammond, Michael"
]
| {P}y{F}oma: a Python finite-state compiler module | acl-demos.24 | Poster | 1807.05252v1 |
https://aclanthology.org/2024.acl-demos.25.bib | @inproceedings{niu-etal-2024-veract,
title = "{V}era{CT} Scan: Retrieval-Augmented Fake News Detection with Justifiable Reasoning",
author = "Niu, Cheng and
Guan, Yang and
Wu, Yuanhao and
Zhu, Juno and
Song, Juntong and
Zhong, Randy and
Zhu, Kaihua and
Xu, Siliang and
Diao, Shizhe and
Zhang, Tong",
editor = "Cao, Yixin and
Feng, Yang and
Xiong, Deyi",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 3: System Demonstrations)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-demos.25",
pages = "266--277",
abstract = "The proliferation of fake news poses a significant threat not only by disseminating misleading information but also by undermining the very foundations of democracy. The recent advance of generative artificial intelligence has further exacerbated the challenge of distinguishing genuine news from fabricated stories. In response to this challenge, we introduce VeraCT Scan, a novel retrieval-augmented system for fake news detection. This system operates by extracting the core facts from a given piece of news and subsequently conducting an internet-wide search to identify corroborating or conflicting reports. Then sources{'} credibility is leveraged for information verification. Besides determining the veracity of news, we also provide transparent evidence and reasoning to support its conclusions, resulting in the interpretability and trust in the results. In addition to GPT-4 Turbo, Llama-2 13B is also fine-tuned for news content understanding, information verification, and reasoning. Both implementations have demonstrated state-of-the-art accuracy in the realm of fake news detection.",
}
| The proliferation of fake news poses a significant threat not only by disseminating misleading information but also by undermining the very foundations of democracy. The recent advance of generative artificial intelligence has further exacerbated the challenge of distinguishing genuine news from fabricated stories. In response to this challenge, we introduce VeraCT Scan, a novel retrieval-augmented system for fake news detection. This system operates by extracting the core facts from a given piece of news and subsequently conducting an internet-wide search to identify corroborating or conflicting reports. Then sources{'} credibility is leveraged for information verification. Besides determining the veracity of news, we also provide transparent evidence and reasoning to support its conclusions, resulting in the interpretability and trust in the results. In addition to GPT-4 Turbo, Llama-2 13B is also fine-tuned for news content understanding, information verification, and reasoning. Both implementations have demonstrated state-of-the-art accuracy in the realm of fake news detection. | [
"Niu, Cheng",
"Guan, Yang",
"Wu, Yuanhao",
"Zhu, Juno",
"Song, Juntong",
"Zhong, R",
"y",
"Zhu, Kaihua",
"Xu, Siliang",
"Diao, Shizhe",
"Zhang, Tong"
]
| {V}era{CT} Scan: Retrieval-Augmented Fake News Detection with Justifiable Reasoning | acl-demos.25 | Poster | 2406.10289v2 |
https://aclanthology.org/2024.acl-demos.26.bib | @inproceedings{suzgun-etal-2024-string2string,
title = "string2string: A Modern Python Library for String-to-String Algorithms",
author = "Suzgun, Mirac and
Shieber, Stuart and
Jurafsky, Dan",
editor = "Cao, Yixin and
Feng, Yang and
Xiong, Deyi",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 3: System Demonstrations)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-demos.26",
pages = "278--285",
abstract = "We introduce **string2string**, an open-source library that offers a comprehensive suite of efficient algorithms for a broad range of string-to-string problems. It includes traditional algorithmic solutions as well as recent advanced neural approaches to tackle various problems in string alignment, distance measurement, lexical and semantic search, and similarity analysis�along with several helpful visualization tools and metrics to facilitate the interpretation and analysis of these methods. Notable algorithms featured in the library include the Smith-Waterman algorithm for pairwise local alignment, the Hirschberg algorithm for global alignment, the Wagner-Fischer algorithm for edit distance, BARTScore and BERTScore for similarity analysis, the Knuth-Morris-Pratt algorithm for lexical search, and Faiss for semantic search. In addition, it wraps existing efficient and widely-used implementations of certain frameworks and metrics, such as sacreBLEU and ROUGE. Overall, the library aims to provide extensive coverage and increased flexibility in comparison to existing libraries for strings. It can be used for many downstream applications, tasks, and problems in natural-language processing, bioinformatics, and computational social sciences. It is implemented in Python, easily installable via pip, and accessible through a simple API. Source code, documentation, and tutorials are all available on our GitHub page: https://github.com/stanfordnlp/string2string* Documentation: https://string2string.readthedocs.io/en/latest/* GitHub page: https://github.com/stanfordnlp/string2string* Short video: https://drive.google.com/file/d/1IT-pBACDVUoEHewk{\_}{\_}5Pz5mU5oAMq5k{\_}/view?usp=sharing",
}
| We introduce **string2string**, an open-source library that offers a comprehensive suite of efficient algorithms for a broad range of string-to-string problems. It includes traditional algorithmic solutions as well as recent advanced neural approaches to tackle various problems in string alignment, distance measurement, lexical and semantic search, and similarity analysis�along with several helpful visualization tools and metrics to facilitate the interpretation and analysis of these methods. Notable algorithms featured in the library include the Smith-Waterman algorithm for pairwise local alignment, the Hirschberg algorithm for global alignment, the Wagner-Fischer algorithm for edit distance, BARTScore and BERTScore for similarity analysis, the Knuth-Morris-Pratt algorithm for lexical search, and Faiss for semantic search. In addition, it wraps existing efficient and widely-used implementations of certain frameworks and metrics, such as sacreBLEU and ROUGE. Overall, the library aims to provide extensive coverage and increased flexibility in comparison to existing libraries for strings. It can be used for many downstream applications, tasks, and problems in natural-language processing, bioinformatics, and computational social sciences. It is implemented in Python, easily installable via pip, and accessible through a simple API. Source code, documentation, and tutorials are all available on our GitHub page: https://github.com/stanfordnlp/string2string* Documentation: https://string2string.readthedocs.io/en/latest/* GitHub page: https://github.com/stanfordnlp/string2string* Short video: https://drive.google.com/file/d/1IT-pBACDVUoEHewk{\_}{\_}5Pz5mU5oAMq5k{\_}/view?usp=sharing | [
"Suzgun, Mirac",
"Shieber, Stuart",
"Jurafsky, Dan"
]
| string2string: A Modern Python Library for String-to-String Algorithms | acl-demos.26 | Poster | 2107.00064v1 |
https://aclanthology.org/2024.acl-demos.27.bib | @inproceedings{liu-etal-2024-proofread,
title = "Proofread: Fixes All Errors with One Tap",
author = "Liu, Renjie and
Zhang, Yanxiang and
Zhu, Yun and
Sun, Haicheng and
Zhang, Yuanbo and
Huang, Michael and
Cai, Shanqing and
Meng, Lei and
Zhai, Shumin",
editor = "Cao, Yixin and
Feng, Yang and
Xiong, Deyi",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 3: System Demonstrations)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-demos.27",
pages = "286--293",
abstract = "The impressive capabilities in Large Language Models (LLMs) provide a powerful approach to reimagine users{'} typing experience. This paper demonstrates the Proofread feature in Gboard, a virtual keyboard running on mobile phones. Proofread enables seamless sentence-level and paragraph-level corrections with a single tap. We describe the complete system in this paper, from data generation, metrics design to model tuning and deployment. To obtain models with sufficient quality, we implement a careful data synthetic pipeline tailored to online use cases, design multifaceted metrics, employ a two-stage tuning approach to acquire the dedicated LLM for the feature: the Supervised Fine Tuning (SFT) for foundational quality, followed by the Reinforcement Learning (RL) tuning approach for targeted refinement. Specifically, we find sequential tuning on Rewrite and proofread tasks yields the best quality in SFT stage, and propose global and direct rewards in the RL tuning stage to seek further improvement. Extensive experiments on a human-labeled golden set showed our tuned PaLM2-XS model achieved 85.56{\%} good ratio. We launched the feature to Pixel 8 devices by serving the model on TPU v5 in Google Cloud, with thousands of daily active users. Serving latency was significantly reduced by quantization, bucket inference, text segmentation, and speculative decoding. Our demo could be seen in Youtube.",
}
| The impressive capabilities in Large Language Models (LLMs) provide a powerful approach to reimagine users{'} typing experience. This paper demonstrates the Proofread feature in Gboard, a virtual keyboard running on mobile phones. Proofread enables seamless sentence-level and paragraph-level corrections with a single tap. We describe the complete system in this paper, from data generation, metrics design to model tuning and deployment. To obtain models with sufficient quality, we implement a careful data synthetic pipeline tailored to online use cases, design multifaceted metrics, employ a two-stage tuning approach to acquire the dedicated LLM for the feature: the Supervised Fine Tuning (SFT) for foundational quality, followed by the Reinforcement Learning (RL) tuning approach for targeted refinement. Specifically, we find sequential tuning on Rewrite and proofread tasks yields the best quality in SFT stage, and propose global and direct rewards in the RL tuning stage to seek further improvement. Extensive experiments on a human-labeled golden set showed our tuned PaLM2-XS model achieved 85.56{\%} good ratio. We launched the feature to Pixel 8 devices by serving the model on TPU v5 in Google Cloud, with thousands of daily active users. Serving latency was significantly reduced by quantization, bucket inference, text segmentation, and speculative decoding. Our demo could be seen in Youtube. | [
"Liu, Renjie",
"Zhang, Yanxiang",
"Zhu, Yun",
"Sun, Haicheng",
"Zhang, Yuanbo",
"Huang, Michael",
"Cai, Shanqing",
"Meng, Lei",
"Zhai, Shumin"
]
| Proofread: Fixes All Errors with One Tap | acl-demos.27 | Poster | 1703.03384v1 |
https://aclanthology.org/2024.acl-demos.28.bib | @inproceedings{nguyen-etal-2024-seallms,
title = "{S}ea{LLM}s - Large Language Models for {S}outheast {A}sia",
author = "Nguyen, Xuan-Phi and
Zhang, Wenxuan and
Li, Xin and
Aljunied, Mahani and
Hu, Zhiqiang and
Shen, Chenhui and
Chia, Yew Ken and
Li, Xingxuan and
Wang, Jianyu and
Tan, Qingyu and
Cheng, Liying and
Chen, Guanzheng and
Deng, Yue and
Yang, Sen and
Liu, Chaoqun and
Zhang, Hang and
Bing, Lidong",
editor = "Cao, Yixin and
Feng, Yang and
Xiong, Deyi",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 3: System Demonstrations)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-demos.28",
pages = "294--304",
abstract = "Despite the remarkable achievements of large language models (LLMs) in various tasks, there remains a linguistic bias that favors high-resource languages, such as English, often at the expense of low-resource and regional languages. To address this imbalance, we introduce SeaLLMs, an innovative series of language models that specifically focuses on Southeast Asian (SEA) languages. SeaLLMs are built upon popular English-centric models through continued pre-training with an extended vocabulary, specialized instruction and alignment tuning to better capture the intricacies of regional languages. This allows them to respect and reflect local cultural norms, customs, stylistic preferences, and legal considerations. Our comprehensive evaluation demonstrates that SeaLLM models exhibit superior performance across a wide spectrum of linguistic tasks and assistant-style instruction-following capabilities relative to comparable open-source models. Moreover, they outperform ChatGPT-3.5 in non-Latin languages, such as Thai, Khmer, Lao, and Burmese, by large margins while remaining lightweight and cost-effective to operate.",
}
| Despite the remarkable achievements of large language models (LLMs) in various tasks, there remains a linguistic bias that favors high-resource languages, such as English, often at the expense of low-resource and regional languages. To address this imbalance, we introduce SeaLLMs, an innovative series of language models that specifically focuses on Southeast Asian (SEA) languages. SeaLLMs are built upon popular English-centric models through continued pre-training with an extended vocabulary, specialized instruction and alignment tuning to better capture the intricacies of regional languages. This allows them to respect and reflect local cultural norms, customs, stylistic preferences, and legal considerations. Our comprehensive evaluation demonstrates that SeaLLM models exhibit superior performance across a wide spectrum of linguistic tasks and assistant-style instruction-following capabilities relative to comparable open-source models. Moreover, they outperform ChatGPT-3.5 in non-Latin languages, such as Thai, Khmer, Lao, and Burmese, by large margins while remaining lightweight and cost-effective to operate. | [
"Nguyen, Xuan-Phi",
"Zhang, Wenxuan",
"Li, Xin",
"Aljunied, Mahani",
"Hu, Zhiqiang",
"Shen, Chenhui",
"Chia, Yew Ken",
"Li, Xingxuan",
"Wang, Jianyu",
"Tan, Qingyu",
"Cheng, Liying",
"Chen, Guanzheng",
"Deng, Yue",
"Yang, Sen",
"Liu, Chaoqun",
"Zhang, Hang",
"Bing, Lidong"
]
| {S}ea{LLM}s - Large Language Models for {S}outheast {A}sia | acl-demos.28 | Poster | 2401.08537v1 |
https://aclanthology.org/2024.acl-demos.29.bib | @inproceedings{dallabetta-etal-2024-fundus,
title = "Fundus: A Simple-to-Use News Scraper Optimized for High Quality Extractions",
author = "Dallabetta, Max and
Dobberstein, Conrad and
Breiding, Adrian and
Akbik, Alan",
editor = "Cao, Yixin and
Feng, Yang and
Xiong, Deyi",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 3: System Demonstrations)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-demos.29",
pages = "305--314",
abstract = "This paper introduces Fundus, a user-friendly news scraper that enables users to obtain millions of high-quality news articles with just a few lines of code. Unlike existing news scrapers, we use manually crafted, bespoke content extractors that are specifically tailored to the formatting guidelines of each supported online newspaper. This allows us to optimize our scraping for quality such that retrieved news articles are textually complete and without HTML artifacts. Further, our framework combines both crawling (retrieving HTML from the web or large web archives) and content extraction into a single pipeline. By providing a unified interface for a predefined collection of newspapers, we aim to make Fundus broadly usable even for non-technical users. This paper gives an overview of the framework, discusses our design choices, and presents a comparative evaluation against other popular news scrapers. Our evaluation shows that Fundus yields significantly higher quality extractions (complete and artifact-free news articles) than prior work.The framework is available on GitHub under https://github.com/flairNLP/fundus and can be simply installed using pip.",
}
| This paper introduces Fundus, a user-friendly news scraper that enables users to obtain millions of high-quality news articles with just a few lines of code. Unlike existing news scrapers, we use manually crafted, bespoke content extractors that are specifically tailored to the formatting guidelines of each supported online newspaper. This allows us to optimize our scraping for quality such that retrieved news articles are textually complete and without HTML artifacts. Further, our framework combines both crawling (retrieving HTML from the web or large web archives) and content extraction into a single pipeline. By providing a unified interface for a predefined collection of newspapers, we aim to make Fundus broadly usable even for non-technical users. This paper gives an overview of the framework, discusses our design choices, and presents a comparative evaluation against other popular news scrapers. Our evaluation shows that Fundus yields significantly higher quality extractions (complete and artifact-free news articles) than prior work.The framework is available on GitHub under https://github.com/flairNLP/fundus and can be simply installed using pip. | [
"Dallabetta, Max",
"Dobberstein, Conrad",
"Breiding, Adrian",
"Akbik, Alan"
]
| Fundus: A Simple-to-Use News Scraper Optimized for High Quality Extractions | acl-demos.29 | Poster | 2403.15279v2 |
https://aclanthology.org/2024.acl-demos.30.bib | @inproceedings{yu-etal-2024-charpoet,
title = "{C}har{P}oet: A {C}hinese Classical Poetry Generation System Based on Token-free {LLM}",
author = "Yu, Chengyue and
Zang, Lei and
Wang, Jiaotuan and
Zhuang, Chenyi and
Gu, Jinjie",
editor = "Cao, Yixin and
Feng, Yang and
Xiong, Deyi",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 3: System Demonstrations)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-demos.30",
pages = "315--325",
abstract = "Automatic Chinese classical poetry generation has attracted much research interest, but achieving effective control over format and content simultaneously remains challenging. Traditional systems usually accept keywords as user inputs, resulting in limited control over content. Large language models (LLMs) improve content control by allowing unrestricted user instructions, but the token-by-token generation process frequently makes format errors. Motivated by this, we propose CharPoet, a Chinese classical poetry generation system based on token-free LLM, which provides effective control over both format and content. Our token-free architecture generates in a character-by-character manner, enabling precise control over the number of characters. Pruned from existing token-based LLMs, CharPoet inherits their pretrained capabilities and can generate poetry following instructions like �Write me a poem for my mother{'}s birthday.� CharPoet achieves format accuracy above 0.96, outperforming Jiuge-GPT-2 (0.91) and GPT-4 (0.38). In terms of content quality, CharPoet surpasses traditional systems including Jiuge, and is comparable to other LLMs. Our system is open source and available at https://modelscope.cn/models/CharPoet/CharPoet. A video demonstration of CharPoet is available at https://youtu.be/voZ25qEp3Dc.",
}
| Automatic Chinese classical poetry generation has attracted much research interest, but achieving effective control over format and content simultaneously remains challenging. Traditional systems usually accept keywords as user inputs, resulting in limited control over content. Large language models (LLMs) improve content control by allowing unrestricted user instructions, but the token-by-token generation process frequently makes format errors. Motivated by this, we propose CharPoet, a Chinese classical poetry generation system based on token-free LLM, which provides effective control over both format and content. Our token-free architecture generates in a character-by-character manner, enabling precise control over the number of characters. Pruned from existing token-based LLMs, CharPoet inherits their pretrained capabilities and can generate poetry following instructions like �Write me a poem for my mother{'}s birthday.� CharPoet achieves format accuracy above 0.96, outperforming Jiuge-GPT-2 (0.91) and GPT-4 (0.38). In terms of content quality, CharPoet surpasses traditional systems including Jiuge, and is comparable to other LLMs. Our system is open source and available at https://modelscope.cn/models/CharPoet/CharPoet. A video demonstration of CharPoet is available at https://youtu.be/voZ25qEp3Dc. | [
"Yu, Chengyue",
"Zang, Lei",
"Wang, Jiaotuan",
"Zhuang, Chenyi",
"Gu, Jinjie"
]
| {C}har{P}oet: A {C}hinese Classical Poetry Generation System Based on Token-free {LLM} | acl-demos.30 | Poster | 1911.08212v1 |
https://aclanthology.org/2024.acl-demos.31.bib | @inproceedings{song-etal-2024-itake,
title = "{ITAKE}: Interactive Unstructured Text Annotation and Knowledge Extraction System with {LLM}s and {M}odel{O}ps",
author = "Song, Jiahe and
Ding, Hongxin and
Wang, Zhiyuan and
Xu, Yongxin and
Wang, Yasha and
Zhao, Junfeng",
editor = "Cao, Yixin and
Feng, Yang and
Xiong, Deyi",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 3: System Demonstrations)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-demos.31",
pages = "326--334",
abstract = "Extracting structured knowledge from unstructured text data has a wide range of application prospects, and a pervasive trend is to develop text annotation tools to help extraction. However, they often encounter issues such as single scenario usage, lack of effective human-machine collaboration, insufficient model supervision, and suboptimal utilization of Large Language Models (LLMs). We introduces an interactive unstructured text annotation and knowledge extraction system that synergistically integrates LLMs and ModelOps to alleviate these issues. The system leverages LLMs for enhanced performance in low-resource contexts, employs a ModelOps platform to monitor models throughout their lifecycle, and amalgamates interactive annotation methods with online machine learning and active learning. The demo video and website are now publicly available.",
}
| Extracting structured knowledge from unstructured text data has a wide range of application prospects, and a pervasive trend is to develop text annotation tools to help extraction. However, they often encounter issues such as single scenario usage, lack of effective human-machine collaboration, insufficient model supervision, and suboptimal utilization of Large Language Models (LLMs). We introduces an interactive unstructured text annotation and knowledge extraction system that synergistically integrates LLMs and ModelOps to alleviate these issues. The system leverages LLMs for enhanced performance in low-resource contexts, employs a ModelOps platform to monitor models throughout their lifecycle, and amalgamates interactive annotation methods with online machine learning and active learning. The demo video and website are now publicly available. | [
"Song, Jiahe",
"Ding, Hongxin",
"Wang, Zhiyuan",
"Xu, Yongxin",
"Wang, Yasha",
"Zhao, Junfeng"
]
| {ITAKE}: Interactive Unstructured Text Annotation and Knowledge Extraction System with {LLM}s and {M}odel{O}ps | acl-demos.31 | Poster | 2310.02593v1 |
https://aclanthology.org/2024.acl-demos.32.bib | @inproceedings{cheng-etal-2024-legent,
title = "{LEGENT}: Open Platform for Embodied Agents",
author = "Cheng, Zhili and
Wang, Zhitong and
Hu, Jinyi and
Hu, Shengding and
Liu, An and
Tu, Yuge and
Li, Pengkai and
Shi, Lei and
Liu, Zhiyuan and
Sun, Maosong",
editor = "Cao, Yixin and
Feng, Yang and
Xiong, Deyi",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 3: System Demonstrations)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-demos.32",
pages = "335--345",
abstract = "Despite advancements in Large Language Models (LLMs) and Large Multimodal Models (LMMs), their integration into language-grounded, human-like embodied agents remains incomplete, hindering complex real-life task performance in 3D environments. Existing integrations often feature limited open-sourcing, challenging collective progress in this field. We introduce LEGENT, an open, scalable platform for developing embodied agents using LLMs and LMMs. LEGENT offers a dual approach: a rich 3D environment with interactive, communicable, and actionable agents, paired with a user-friendly interface, and a sophisticated data generation pipeline utilizing advanced algorithms to exploit supervision from simulated worlds at scale. In our experiments, an embryonic vision-language-action model trained on LEGENT-generated data surpasses GPT-4V in embodied tasks, showcasing promising generalization capabilities. The demo video is available at the following link https://video.legent.ai.",
}
| Despite advancements in Large Language Models (LLMs) and Large Multimodal Models (LMMs), their integration into language-grounded, human-like embodied agents remains incomplete, hindering complex real-life task performance in 3D environments. Existing integrations often feature limited open-sourcing, challenging collective progress in this field. We introduce LEGENT, an open, scalable platform for developing embodied agents using LLMs and LMMs. LEGENT offers a dual approach: a rich 3D environment with interactive, communicable, and actionable agents, paired with a user-friendly interface, and a sophisticated data generation pipeline utilizing advanced algorithms to exploit supervision from simulated worlds at scale. In our experiments, an embryonic vision-language-action model trained on LEGENT-generated data surpasses GPT-4V in embodied tasks, showcasing promising generalization capabilities. The demo video is available at the following link https://video.legent.ai. | [
"Cheng, Zhili",
"Wang, Zhitong",
"Hu, Jinyi",
"Hu, Shengding",
"Liu, An",
"Tu, Yuge",
"Li, Pengkai",
"Shi, Lei",
"Liu, Zhiyuan",
"Sun, Maosong"
]
| {LEGENT}: Open Platform for Embodied Agents | acl-demos.32 | Poster | 1904.01201v2 |
https://aclanthology.org/2024.acl-demos.33.bib | @inproceedings{ramponi-etal-2024-variationist,
title = "Variationist: Exploring Multifaceted Variation and Bias in Written Language Data",
author = "Ramponi, Alan and
Casula, Camilla and
Menini, Stefano",
editor = "Cao, Yixin and
Feng, Yang and
Xiong, Deyi",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 3: System Demonstrations)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-demos.33",
pages = "346--354",
abstract = "Exploring and understanding language data is a fundamental stage in all areas dealing with human language. It allows NLP practitioners to uncover quality concerns and harmful biases in data before training, and helps linguists and social scientists to gain insight into language use and human behavior. Yet, there is currently a lack of a unified, customizable tool to seamlessly inspect and visualize language variation and bias across multiple variables, language units, and diverse metrics that go beyond descriptive statistics. In this paper, we introduce Variationist, a highly-modular, extensible, and task-agnostic tool that fills this gap. Variationist handles at once a potentially unlimited combination of variable types and semantics across diversity and association metrics with regards to the language unit of choice, and orchestrates the creation of up to five-dimensional interactive charts for over 30 variable type-semantics combinations. Through our case studies on computational dialectology, human label variation, and text generation, we show how Variationist enables researchers from different disciplines to effortlessly answer specific research questions or unveil undesired associations in language data. A Python library, code, documentation, and tutorials are made publicly available to the research community.",
}
| Exploring and understanding language data is a fundamental stage in all areas dealing with human language. It allows NLP practitioners to uncover quality concerns and harmful biases in data before training, and helps linguists and social scientists to gain insight into language use and human behavior. Yet, there is currently a lack of a unified, customizable tool to seamlessly inspect and visualize language variation and bias across multiple variables, language units, and diverse metrics that go beyond descriptive statistics. In this paper, we introduce Variationist, a highly-modular, extensible, and task-agnostic tool that fills this gap. Variationist handles at once a potentially unlimited combination of variable types and semantics across diversity and association metrics with regards to the language unit of choice, and orchestrates the creation of up to five-dimensional interactive charts for over 30 variable type-semantics combinations. Through our case studies on computational dialectology, human label variation, and text generation, we show how Variationist enables researchers from different disciplines to effortlessly answer specific research questions or unveil undesired associations in language data. A Python library, code, documentation, and tutorials are made publicly available to the research community. | [
"Ramponi, Alan",
"Casula, Camilla",
"Menini, Stefano"
]
| Variationist: Exploring Multifaceted Variation and Bias in Written Language Data | acl-demos.33 | Poster | 2406.17647v1 |
https://aclanthology.org/2024.acl-demos.34.bib | @inproceedings{wysocki-etal-2024-llm,
title = "An {LLM}-based Knowledge Synthesis and Scientific Reasoning Framework for Biomedical Discovery",
author = "Wysocki, Oskar and
[email protected], [email protected] and
Carvalho, Danilo and
Bogatu, Alex and
[email protected], [email protected] and
[email protected], [email protected] and
[email protected], [email protected] and
Freitas, Andre",
editor = "Cao, Yixin and
Feng, Yang and
Xiong, Deyi",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 3: System Demonstrations)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-demos.34",
pages = "355--364",
abstract = "We present BioLunar, developed using the Lunar framework, as a tool for supporting biological analyses, with a particular emphasis on molecular-level evidence enrichment for biomarker discovery in oncology. The platform integrates Large Language Models (LLMs) to facilitate complex scientific reasoning across distributed evidence spaces, enhancing the capability for harmonizing and reasoning over heterogeneous data sources. Demonstrating its utility in cancer research, BioLunar leverages modular design, reusable data access and data analysis components, and a low-code user interface, enabling researchers of all programming levels to construct LLM-enabled scientific workflows. By facilitating automatic scientific discovery and inference from heterogeneous evidence, BioLunar exemplifies the potential of the integration between LLMs, specialised databases and biomedical tools to support expert-level knowledge synthesis and discovery.",
}
| We present BioLunar, developed using the Lunar framework, as a tool for supporting biological analyses, with a particular emphasis on molecular-level evidence enrichment for biomarker discovery in oncology. The platform integrates Large Language Models (LLMs) to facilitate complex scientific reasoning across distributed evidence spaces, enhancing the capability for harmonizing and reasoning over heterogeneous data sources. Demonstrating its utility in cancer research, BioLunar leverages modular design, reusable data access and data analysis components, and a low-code user interface, enabling researchers of all programming levels to construct LLM-enabled scientific workflows. By facilitating automatic scientific discovery and inference from heterogeneous evidence, BioLunar exemplifies the potential of the integration between LLMs, specialised databases and biomedical tools to support expert-level knowledge synthesis and discovery. | [
"Wysocki, Oskar",
"[email protected], [email protected]",
"Carvalho, Danilo",
"Bogatu, Alex",
"Danilo.mir",
"[email protected], Danilo.mir",
"[email protected]",
"[email protected], [email protected]",
"[email protected], [email protected]",
"Freitas, Andre"
]
| An {LLM}-based Knowledge Synthesis and Scientific Reasoning Framework for Biomedical Discovery | acl-demos.34 | Poster | 2406.18626v1 |
https://aclanthology.org/2024.acl-demos.35.bib | @inproceedings{zhou-etal-2024-cogmg,
title = "{C}og{MG}: Collaborative Augmentation Between Large Language Model and Knowledge Graph",
author = "Zhou, Tong and
Chen, Yubo and
Liu, Kang and
Zhao, Jun",
editor = "Cao, Yixin and
Feng, Yang and
Xiong, Deyi",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 3: System Demonstrations)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-demos.35",
pages = "365--373",
abstract = "Large language models have become integral to question-answering applications despite their propensity for generating hallucinations and factually inaccurate content. Querying knowledge graphs to reduce hallucinations in LLM meets the challenge of incomplete knowledge coverage in knowledge graphs. On the other hand, updating knowledge graphs by information extraction and knowledge graph completion faces the knowledge update misalignment issue. In this work, we introduce a collaborative augmentation framework, CogMG, leveraging knowledge graphs to address the limitations of LLMs in QA scenarios, explicitly targeting the problems of incomplete knowledge coverage and knowledge update misalignment. The LLMs identify and decompose required knowledge triples that are not present in the KG, enriching them and aligning updates with real-world demands. We demonstrate the efficacy of this approach through a supervised fine-tuned LLM within an agent framework, showing significant improvements in reducing hallucinations and enhancing factual accuracy in QA responses. Our code and video are publicly available.",
}
| Large language models have become integral to question-answering applications despite their propensity for generating hallucinations and factually inaccurate content. Querying knowledge graphs to reduce hallucinations in LLM meets the challenge of incomplete knowledge coverage in knowledge graphs. On the other hand, updating knowledge graphs by information extraction and knowledge graph completion faces the knowledge update misalignment issue. In this work, we introduce a collaborative augmentation framework, CogMG, leveraging knowledge graphs to address the limitations of LLMs in QA scenarios, explicitly targeting the problems of incomplete knowledge coverage and knowledge update misalignment. The LLMs identify and decompose required knowledge triples that are not present in the KG, enriching them and aligning updates with real-world demands. We demonstrate the efficacy of this approach through a supervised fine-tuned LLM within an agent framework, showing significant improvements in reducing hallucinations and enhancing factual accuracy in QA responses. Our code and video are publicly available. | [
"Zhou, Tong",
"Chen, Yubo",
"Liu, Kang",
"Zhao, Jun"
]
| {C}og{MG}: Collaborative Augmentation Between Large Language Model and Knowledge Graph | acl-demos.35 | Poster | 2306.04136v1 |
https://aclanthology.org/2024.acl-demos.36.bib | @inproceedings{hu-etal-2024-ella,
title = "{ELLA}: Empowering {LLM}s for Interpretable, Accurate and Informative Legal Advice",
author = "Hu, Yutong and
Luo, Kangcheng and
Feng, Yansong",
editor = "Cao, Yixin and
Feng, Yang and
Xiong, Deyi",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 3: System Demonstrations)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-demos.36",
pages = "374--387",
abstract = "Despite remarkable performance in legal consultation exhibited by legal Large Language Models(LLMs) combined with legal article retrieval components, there are still cases when the advice given is incorrect or baseless. To alleviate these problems, we propose \textbf{ELLA}, a tool for \textbf{E}mpowering \textbf{L}LMs for interpretable, accurate, and informative \textbf{L}egal \textbf{A}dvice. ELLA visually presents the correlation between legal articles and LLM{'}s response by calculating their similarities, providing users with an intuitive legal basis for the responses. Besides, based on the users{'} queries, ELLA retrieves relevant legal articles and displays them to users. Users can interactively select legal articles for LLM to generate more accurate responses. ELLA also retrieves relevant legal cases for user reference. Our user study shows that presenting the legal basis for the response helps users understand better. The accuracy of LLM{'}s responses also improves when users intervene in selecting legal articles for LLM. Providing relevant legal cases also aids individuals in obtaining comprehensive information. Our github repo is: \url{https://github.com/Huyt00/ELLA}.",
}
| Despite remarkable performance in legal consultation exhibited by legal Large Language Models(LLMs) combined with legal article retrieval components, there are still cases when the advice given is incorrect or baseless. To alleviate these problems, we propose \textbf{ELLA}, a tool for \textbf{E}mpowering \textbf{L}LMs for interpretable, accurate, and informative \textbf{L}egal \textbf{A}dvice. ELLA visually presents the correlation between legal articles and LLM{'}s response by calculating their similarities, providing users with an intuitive legal basis for the responses. Besides, based on the users{'} queries, ELLA retrieves relevant legal articles and displays them to users. Users can interactively select legal articles for LLM to generate more accurate responses. ELLA also retrieves relevant legal cases for user reference. Our user study shows that presenting the legal basis for the response helps users understand better. The accuracy of LLM{'}s responses also improves when users intervene in selecting legal articles for LLM. Providing relevant legal cases also aids individuals in obtaining comprehensive information. Our github repo is: \url{https://github.com/Huyt00/ELLA}. | [
"Hu, Yutong",
"Luo, Kangcheng",
"Feng, Yansong"
]
| {ELLA}: Empowering {LLM}s for Interpretable, Accurate and Informative Legal Advice | acl-demos.36 | Poster | 2402.01864v2 |
https://aclanthology.org/2024.acl-demos.37.bib | @inproceedings{tang-etal-2024-llmbox,
title = "{LLMB}ox: A Comprehensive Library for Large Language Models",
author = "Tang, Tianyi and
Yiwen, Hu and
Li, Bingqian and
Luo, Wenyang and
Qin, ZiJing and
Sun, Haoxiang and
Wang, Jiapeng and
Xu, Shiyi and
Cheng, Xiaoxue and
Guo, Geyang and
Peng, Han and
Zheng, Bowen and
Tang, Yiru and
Min, Yingqian and
Chen, Yushuo and
Chen, Jie and
Zhao, Ranchi and
Ding, Luran and
Wang, Yuhao and
Dong, Zican and
Chunxuan, Xia and
Li, Junyi and
Zhou, Kun and
Zhao, Xin and
Wen, Ji-Rong",
editor = "Cao, Yixin and
Feng, Yang and
Xiong, Deyi",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 3: System Demonstrations)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-demos.37",
pages = "388--399",
abstract = "To facilitate the research on large language models (LLMs), this paper presents a comprehensive and unified library, LLMBox, to ease the development, use, and evaluation of LLMs. This library is featured with three main merits: (1) a unified data interface that supports the flexible implementation of various training strategies, (2) a comprehensive evaluation that covers extensive tasks, datasets, and models, and (3) more practical consideration, especially on user-friendliness and efficiency. With our library, users can easily reproduce existing methods, train new models, and conduct comprehensive performance comparisons. To rigorously test LLMBox, we conduct extensive experiments in a diverse coverage of evaluation settings, and experimental results demonstrate the effectiveness and efficiency of our library in supporting various implementations related to LLMs. The detailed introduction and usage guidance can be found at \url{https://github.com/RUCAIBox/LLMBox}.",
}
| To facilitate the research on large language models (LLMs), this paper presents a comprehensive and unified library, LLMBox, to ease the development, use, and evaluation of LLMs. This library is featured with three main merits: (1) a unified data interface that supports the flexible implementation of various training strategies, (2) a comprehensive evaluation that covers extensive tasks, datasets, and models, and (3) more practical consideration, especially on user-friendliness and efficiency. With our library, users can easily reproduce existing methods, train new models, and conduct comprehensive performance comparisons. To rigorously test LLMBox, we conduct extensive experiments in a diverse coverage of evaluation settings, and experimental results demonstrate the effectiveness and efficiency of our library in supporting various implementations related to LLMs. The detailed introduction and usage guidance can be found at \url{https://github.com/RUCAIBox/LLMBox}. | [
"Tang, Tianyi",
"Yiwen, Hu",
"Li, Bingqian",
"Luo, Wenyang",
"Qin, ZiJing",
"Sun, Haoxiang",
"Wang, Jiapeng",
"Xu, Shiyi",
"Cheng, Xiaoxue",
"Guo, Geyang",
"Peng, Han",
"Zheng, Bowen",
"Tang, Yiru",
"Min, Yingqian",
"Chen, Yushuo",
"Chen, Jie",
"Zhao, Ranchi",
"Ding, Luran",
"Wang, Yuhao",
"Dong, Zican",
"Chunxuan, Xia",
"Li, Junyi",
"Zhou, Kun",
"Zhao, Xin",
"Wen, Ji-Rong"
]
| {LLMB}ox: A Comprehensive Library for Large Language Models | acl-demos.37 | Poster | 2311.09635v2 |
https://aclanthology.org/2024.acl-demos.38.bib | @inproceedings{zheng-etal-2024-llamafactory,
title = "{L}lama{F}actory: Unified Efficient Fine-Tuning of 100+ Language Models",
author = "Zheng, Yaowei and
Zhang, Richong and
Zhang, Junhao and
YeYanhan, YeYanhan and
Luo, Zheyan",
editor = "Cao, Yixin and
Feng, Yang and
Xiong, Deyi",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 3: System Demonstrations)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-demos.38",
pages = "400--410",
abstract = "Efficient fine-tuning is vital for adapting large language models (LLMs) to downstream tasks. However, it requires non-trivial efforts to implement these methods on different models. We present LlamaFactory, a unified framework that integrates a suite of cutting-edge efficient training methods. It provides a solution for flexibly customizing the fine-tuning of 100+ LLMs without the need for coding through the built-in web UI LlamaBoard. We empirically validate the efficiency and effectiveness of our framework on language modeling and text generation tasks. It has been released at https://github.com/hiyouga/LLaMA-Factory and received over 25,000 stars and 3,000 forks.",
}
| Efficient fine-tuning is vital for adapting large language models (LLMs) to downstream tasks. However, it requires non-trivial efforts to implement these methods on different models. We present LlamaFactory, a unified framework that integrates a suite of cutting-edge efficient training methods. It provides a solution for flexibly customizing the fine-tuning of 100+ LLMs without the need for coding through the built-in web UI LlamaBoard. We empirically validate the efficiency and effectiveness of our framework on language modeling and text generation tasks. It has been released at https://github.com/hiyouga/LLaMA-Factory and received over 25,000 stars and 3,000 forks. | [
"Zheng, Yaowei",
"Zhang, Richong",
"Zhang, Junhao",
"YeYanhan, YeYanhan",
"Luo, Zheyan"
]
| {L}lama{F}actory: Unified Efficient Fine-Tuning of 100+ Language Models | acl-demos.38 | Poster | 2403.13372v4 |
https://aclanthology.org/2024.acl-srw.1.bib | @inproceedings{keita-etal-2024-feriji,
title = "Feriji: A {F}rench-{Z}arma Parallel Corpus, Glossary {\&} Translator",
author = "Keita, Mamadou and
Ibrahim, Elysabhete and
Alfari, Habibatou and
Homan, Christopher",
editor = "Fu, Xiyan and
Fleisig, Eve",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 4: Student Research Workshop)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-srw.1",
pages = "1--9",
abstract = "Machine translation (MT) is a rapidly expanding field that has experienced significant advancements in recent years with the development of models capable of translating multiple languages with remarkable accuracy. However, the representation of African languages in this field still needs improvement due to linguistic complexities and limited resources. This applies to the Zarma language, a dialect of Songhay (of the Nilo-Saharan language family) spoken by over 5 million people across Niger and neighboring countries (Lewis et al., 2016). This paper introduces Feriji, the first robust French-Zarma parallel corpus and glossary designed for MT. The corpus, containing 61,085 sentences in Zarma and 42,789 in French, and a glossary of 4,062 words represents a significant step in addressing the need for more resources for Zarma. We fine-tune three large language models on our dataset, obtaining a BLEU score of 30.06 on the best-performing model. We further evaluate the models on human judgments of fluency, comprehension, and readability and the importance and impact of the corpus and models. Our contributions help to bridge a significant language gap and promote an essential and overlooked indigenous African language.",
}
| Machine translation (MT) is a rapidly expanding field that has experienced significant advancements in recent years with the development of models capable of translating multiple languages with remarkable accuracy. However, the representation of African languages in this field still needs improvement due to linguistic complexities and limited resources. This applies to the Zarma language, a dialect of Songhay (of the Nilo-Saharan language family) spoken by over 5 million people across Niger and neighboring countries (Lewis et al., 2016). This paper introduces Feriji, the first robust French-Zarma parallel corpus and glossary designed for MT. The corpus, containing 61,085 sentences in Zarma and 42,789 in French, and a glossary of 4,062 words represents a significant step in addressing the need for more resources for Zarma. We fine-tune three large language models on our dataset, obtaining a BLEU score of 30.06 on the best-performing model. We further evaluate the models on human judgments of fluency, comprehension, and readability and the importance and impact of the corpus and models. Our contributions help to bridge a significant language gap and promote an essential and overlooked indigenous African language. | [
"Keita, Mamadou",
"Ibrahim, Elysabhete",
"Alfari, Habibatou",
"Homan, Christopher"
]
| Feriji: A {F}rench-{Z}arma Parallel Corpus, Glossary {\&} Translator | acl-srw.1 | Poster | 2406.05888v2 |
https://aclanthology.org/2024.acl-srw.2.bib | @inproceedings{cho-kim-2024-pragmatic,
title = "Pragmatic inference of scalar implicature by {LLM}s",
author = "Cho, Ye-eun and
Kim, Seong mook",
editor = "Fu, Xiyan and
Fleisig, Eve",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 4: Student Research Workshop)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-srw.2",
pages = "10--20",
abstract = "This study investigates how Large Language Models (LLMs), particularly BERT (Devlin et al., 2019) and GPT-2 (Radford et al., 2019), engage in pragmatic inference of scalar implicature, such as some. Two sets of experiments were conducted using cosine similarity and next sentence/token prediction as experimental methods. The results in experiment 1 showed that, both models interpret some as pragmatic implicature not all in the absence of context, aligning with human language processing. In experiment 2, in which Question Under Discussion (QUD) was presented as a contextual cue, BERT showed consistent performance regardless of types of QUDs, while GPT-2 encountered processing difficulties since a certain type of QUD required pragmatic inference for implicature. The findings revealed that, in terms of theoretical approaches, BERT inherently incorporates pragmatic implicature not all within the term some, adhering to Default model (Levinson, 2000). In contrast, GPT-2 seems to encounter processing difficulties in inferring pragmatic implicature within context, consistent with Context-driven model (Sperber and Wilson, 2002).",
}
| This study investigates how Large Language Models (LLMs), particularly BERT (Devlin et al., 2019) and GPT-2 (Radford et al., 2019), engage in pragmatic inference of scalar implicature, such as some. Two sets of experiments were conducted using cosine similarity and next sentence/token prediction as experimental methods. The results in experiment 1 showed that, both models interpret some as pragmatic implicature not all in the absence of context, aligning with human language processing. In experiment 2, in which Question Under Discussion (QUD) was presented as a contextual cue, BERT showed consistent performance regardless of types of QUDs, while GPT-2 encountered processing difficulties since a certain type of QUD required pragmatic inference for implicature. The findings revealed that, in terms of theoretical approaches, BERT inherently incorporates pragmatic implicature not all within the term some, adhering to Default model (Levinson, 2000). In contrast, GPT-2 seems to encounter processing difficulties in inferring pragmatic implicature within context, consistent with Context-driven model (Sperber and Wilson, 2002). | [
"Cho, Ye-eun",
"Kim, Seong mook"
]
| Pragmatic inference of scalar implicature by {LLM}s | acl-srw.2 | Poster | 2408.06673v1 |
https://aclanthology.org/2024.acl-srw.3.bib | @inproceedings{doi-etal-2024-topic,
title = "Topic Modeling for Short Texts with Large Language Models",
author = "Doi, Tomoki and
Isonuma, Masaru and
Yanaka, Hitomi",
editor = "Fu, Xiyan and
Fleisig, Eve",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 4: Student Research Workshop)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-srw.3",
pages = "21--33",
abstract = "As conventional topic models rely on word co-occurrence to infer latent topics, topic modeling for short texts has been a long-standing challenge. Large Language Models (LLMs) can potentially overcome this challenge by contextually learning the meanings of words via pretraining. In this paper, we study two approaches to using LLMs for topic modeling: parallel prompting and sequential prompting. Input length limitations prevent LLMs from processing many texts at once. However, an arbitrary number of texts can be handled by LLMs by splitting the texts into smaller subsets and processing them in parallel or sequentially. Our experimental results demonstrate that our methods can identify more coherent topics than existing ones while maintaining the diversity of the induced topics. Furthermore, we found that the inferred topics cover the input texts to some extent, while hallucinated topics are hardly generated.",
}
| As conventional topic models rely on word co-occurrence to infer latent topics, topic modeling for short texts has been a long-standing challenge. Large Language Models (LLMs) can potentially overcome this challenge by contextually learning the meanings of words via pretraining. In this paper, we study two approaches to using LLMs for topic modeling: parallel prompting and sequential prompting. Input length limitations prevent LLMs from processing many texts at once. However, an arbitrary number of texts can be handled by LLMs by splitting the texts into smaller subsets and processing them in parallel or sequentially. Our experimental results demonstrate that our methods can identify more coherent topics than existing ones while maintaining the diversity of the induced topics. Furthermore, we found that the inferred topics cover the input texts to some extent, while hallucinated topics are hardly generated. | [
"Doi, Tomoki",
"Isonuma, Masaru",
"Yanaka, Hitomi"
]
| Topic Modeling for Short Texts with Large Language Models | acl-srw.3 | Poster | 2403.17706v1 |
https://aclanthology.org/2024.acl-srw.4.bib | @inproceedings{zhang-etal-2024-llms,
title = "Can {LLM}s substitute {SQL}? Comparing Resource Utilization of Querying {LLM}s versus Traditional Relational Databases",
author = "Zhang, Xiang and
Khedri, Khatoon and
Rawassizadeh, Reza",
editor = "Fu, Xiyan and
Fleisig, Eve",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 4: Student Research Workshop)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-srw.4",
pages = "34--41",
abstract = "Large Language Models (LLMs) can automate or substitute different types of tasks in the software engineering process. This study evaluates the resource utilization and accuracy of LLM in interpreting and executing natural language queries against traditional SQL within relational database management systems. We empirically examine the resource utilization and accuracy of nine LLMs varying from 7 to 34 Billion parameters, including Llama2 7B, Llama2 13B, Mistral, Mixtral, Optimus-7B, SUS-chat-34B, platypus-yi-34b, NeuralHermes-2.5-Mistral-7B and Starling-LM-7B-alpha, using a small transaction dataset. Our findings indicate that using LLMs for database queries incurs significant energy overhead (even small and quantized models), making it an environmentally unfriendly approach. Therefore, we advise against replacing relational databases with LLMs due to their substantial resource utilization.",
}
| Large Language Models (LLMs) can automate or substitute different types of tasks in the software engineering process. This study evaluates the resource utilization and accuracy of LLM in interpreting and executing natural language queries against traditional SQL within relational database management systems. We empirically examine the resource utilization and accuracy of nine LLMs varying from 7 to 34 Billion parameters, including Llama2 7B, Llama2 13B, Mistral, Mixtral, Optimus-7B, SUS-chat-34B, platypus-yi-34b, NeuralHermes-2.5-Mistral-7B and Starling-LM-7B-alpha, using a small transaction dataset. Our findings indicate that using LLMs for database queries incurs significant energy overhead (even small and quantized models), making it an environmentally unfriendly approach. Therefore, we advise against replacing relational databases with LLMs due to their substantial resource utilization. | [
"Zhang, Xiang",
"Khedri, Khatoon",
"Rawassizadeh, Reza"
]
| Can {LLM}s substitute {SQL}? Comparing Resource Utilization of Querying {LLM}s versus Traditional Relational Databases | acl-srw.4 | Poster | 2404.08727v1 |
https://aclanthology.org/2024.acl-srw.5.bib | @inproceedings{wang-etal-2024-speech,
title = "Speech-to-Speech Translation with Discrete-Unit-Based Style Transfer",
author = "Wang, Yongqi and
Jionghao, Bai and
Huang, Rongjie and
Li, Ruiqi and
Hong, Zhiqing and
Zhao, Zhou",
editor = "Fu, Xiyan and
Fleisig, Eve",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 4: Student Research Workshop)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-srw.5",
pages = "42--49",
abstract = "Direct speech-to-speech translation (S2ST) with discrete self-supervised representations has achieved remarkable accuracy, but is unable to preserve the speaker timbre of the source speech. Meanwhile, the scarcity of high-quality speaker-parallel data poses a challenge for learning style transfer during translation. We design an S2ST pipeline with style-transfer capability on the basis of discrete self-supervised speech representations and codec units. The acoustic language model we introduce for style transfer leverages self-supervised in-context learning, acquiring style transfer ability without relying on any speaker-parallel data, thereby overcoming data scarcity. By using extensive training data, our model achieves zero-shot cross-lingual style transfer on previously unseen source languages. Experiments show that our model generates translated speeches with high fidelity and speaker similarity. Audio samples are available at http://stylelm.github.io/ .",
}
| Direct speech-to-speech translation (S2ST) with discrete self-supervised representations has achieved remarkable accuracy, but is unable to preserve the speaker timbre of the source speech. Meanwhile, the scarcity of high-quality speaker-parallel data poses a challenge for learning style transfer during translation. We design an S2ST pipeline with style-transfer capability on the basis of discrete self-supervised speech representations and codec units. The acoustic language model we introduce for style transfer leverages self-supervised in-context learning, acquiring style transfer ability without relying on any speaker-parallel data, thereby overcoming data scarcity. By using extensive training data, our model achieves zero-shot cross-lingual style transfer on previously unseen source languages. Experiments show that our model generates translated speeches with high fidelity and speaker similarity. Audio samples are available at http://stylelm.github.io/ . | [
"Wang, Yongqi",
"Jionghao, Bai",
"Huang, Rongjie",
"Li, Ruiqi",
"Hong, Zhiqing",
"Zhao, Zhou"
]
| Speech-to-Speech Translation with Discrete-Unit-Based Style Transfer | acl-srw.5 | Poster | 1808.07894v1 |
https://aclanthology.org/2024.acl-srw.6.bib | @inproceedings{li-etal-2024-instructcoder,
title = "{I}nstruct{C}oder: Instruction Tuning Large Language Models for Code Editing",
author = "Li, Kaixin and
Hu, Qisheng and
Zhao, James and
Chen, Hui and
Xie, Yuxi and
Liu, Tiedong and
Shieh, Michael and
He, Junxian",
editor = "Fu, Xiyan and
Fleisig, Eve",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 4: Student Research Workshop)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-srw.6",
pages = "50--70",
abstract = "Code editing encompasses a variety of pragmatic tasks that developers deal with daily. Despite its relevance and practical usefulness, automatic code editing remains an underexplored area in the evolution of deep learning models, partly due to data scarcity. In this work, we explore the use of Large Language Models (LLMs) to edit code based on user instructions. Evaluated on a novel human-written execution-based benchmark dubbed EditEval, we found current models often struggle to fulfill the instructions. In light of this, we contribute InstructCoder, the first instruction-tuning dataset designed to adapt LLMs for general-purpose code editing, containing high-diversity code-editing tasks such as comment insertion, code optimization, and code refactoring. It consists of over 114,000 instruction-input-output triplets and covers multiple distinct code editing scenarios. The collection process starts with filtered commit data sourced from GitHub Python repositories as seeds. Subsequently, the dataset is systematically expanded through an iterative process, where both seed and generated tasks are used to prompt ChatGPT for more data. Our findings reveal that open-source LLMs fine-tuned on InstructCoder can significantly enhance the accuracy of code edits, exhibiting superior code-editing performance matching advanced proprietary LLMs. The datasets and the source code are publicly available.",
}
| Code editing encompasses a variety of pragmatic tasks that developers deal with daily. Despite its relevance and practical usefulness, automatic code editing remains an underexplored area in the evolution of deep learning models, partly due to data scarcity. In this work, we explore the use of Large Language Models (LLMs) to edit code based on user instructions. Evaluated on a novel human-written execution-based benchmark dubbed EditEval, we found current models often struggle to fulfill the instructions. In light of this, we contribute InstructCoder, the first instruction-tuning dataset designed to adapt LLMs for general-purpose code editing, containing high-diversity code-editing tasks such as comment insertion, code optimization, and code refactoring. It consists of over 114,000 instruction-input-output triplets and covers multiple distinct code editing scenarios. The collection process starts with filtered commit data sourced from GitHub Python repositories as seeds. Subsequently, the dataset is systematically expanded through an iterative process, where both seed and generated tasks are used to prompt ChatGPT for more data. Our findings reveal that open-source LLMs fine-tuned on InstructCoder can significantly enhance the accuracy of code edits, exhibiting superior code-editing performance matching advanced proprietary LLMs. The datasets and the source code are publicly available. | [
"Li, Kaixin",
"Hu, Qisheng",
"Zhao, James",
"Chen, Hui",
"Xie, Yuxi",
"Liu, Tiedong",
"Shieh, Michael",
"He, Junxian"
]
| {I}nstruct{C}oder: Instruction Tuning Large Language Models for Code Editing | acl-srw.6 | Poster | 2310.20329v3 |
https://aclanthology.org/2024.acl-srw.7.bib | @inproceedings{allam-2024-biasdpo,
title = "{B}ias{DPO}: Mitigating Bias in Language Models through Direct Preference Optimization",
author = "Allam, Ahmed",
editor = "Fu, Xiyan and
Fleisig, Eve",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 4: Student Research Workshop)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-srw.7",
pages = "71--79",
abstract = "Large Language Models (LLMs) have become pivotal in advancing natural language processing, yet their potential to perpetuate biases poses significant concerns. This paper introduces a new framework employing Direct Preference Optimization (DPO) to mitigate gender, racial, and religious biases in LLM-generated English text. By developing a loss function that favors less biased over biased completions, our approach cultivates a preference for respectful and non-discriminatory language in LLMs. We also contribute a manually designed dataset for training LLMs to recognize and correct biases. This dataset encompasses a diverse range of prompts paired with both biased and unbiased completions. Implementing this approach on the Microsoft Phi-2 model, we demonstrate substantial reductions in biased outputs as our model outperforms the baseline model on almost all bias benchmarks. Our model also achieves better performance compared to other open-source models on most benchmarks. By reducing biases in the language generated by the model, our study marks a significant step towards developing more ethical and socially responsible LLMs. We publicly release BiasDPO dataset on HuggingFace.",
}
| Large Language Models (LLMs) have become pivotal in advancing natural language processing, yet their potential to perpetuate biases poses significant concerns. This paper introduces a new framework employing Direct Preference Optimization (DPO) to mitigate gender, racial, and religious biases in LLM-generated English text. By developing a loss function that favors less biased over biased completions, our approach cultivates a preference for respectful and non-discriminatory language in LLMs. We also contribute a manually designed dataset for training LLMs to recognize and correct biases. This dataset encompasses a diverse range of prompts paired with both biased and unbiased completions. Implementing this approach on the Microsoft Phi-2 model, we demonstrate substantial reductions in biased outputs as our model outperforms the baseline model on almost all bias benchmarks. Our model also achieves better performance compared to other open-source models on most benchmarks. By reducing biases in the language generated by the model, our study marks a significant step towards developing more ethical and socially responsible LLMs. We publicly release BiasDPO dataset on HuggingFace. | [
"Allam, Ahmed"
]
| {B}ias{DPO}: Mitigating Bias in Language Models through Direct Preference Optimization | acl-srw.7 | Poster | 2407.13928v1 |
https://aclanthology.org/2024.acl-srw.8.bib | @inproceedings{zhong-etal-2024-moextend,
title = "{M}o{E}xtend: Tuning New Experts for Modality and Task Extension",
author = "Zhong, Shanshan and
Gao, Shanghua and
Huang, Zhongzhan and
Wen, Wushao and
Zitnik, Marinka and
Zhou, Pan",
editor = "Fu, Xiyan and
Fleisig, Eve",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 4: Student Research Workshop)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-srw.8",
pages = "80--91",
abstract = "Large language models (LLMs) excel in various tasks but are primarily trained on text data, limiting their application scope. Expanding LLM capabilities to include vision-language understanding is vital, yet training them on multimodal data from scratch is challenging and costly. Existing instruction tuning methods, e.g., LLAVA, often connects a pretrained CLIP vision encoder and LLMs via fully fine-tuning LLMs to bridge the modality gap. However, full fine-tuning is plagued by catastrophic forgetting, i.e., forgetting previous knowledge, and high training costs particularly in the era of increasing tasks and modalities. To solve this issue, we introduce MoExtend, an effective framework designed to streamline the modality adaptation and extension of Mixture-of-Experts (MoE) models. MoExtend seamlessly integrates new experts into pre-trained MoE models, endowing them with novel knowledge without the need to tune pretrained models such as MoE and vision encoders. This approach enables rapid adaptation and extension to new modal data or tasks, effectively addressing the challenge of accommodating new modalities within LLMs. Furthermore, MoExtend avoids tuning pretrained models, thus mitigating the risk of catastrophic forgetting. Experimental results demonstrate the efficacy and efficiency of MoExtend in enhancing the multimodal capabilities of LLMs, contributing to advancements in multimodal AI research.",
}
| Large language models (LLMs) excel in various tasks but are primarily trained on text data, limiting their application scope. Expanding LLM capabilities to include vision-language understanding is vital, yet training them on multimodal data from scratch is challenging and costly. Existing instruction tuning methods, e.g., LLAVA, often connects a pretrained CLIP vision encoder and LLMs via fully fine-tuning LLMs to bridge the modality gap. However, full fine-tuning is plagued by catastrophic forgetting, i.e., forgetting previous knowledge, and high training costs particularly in the era of increasing tasks and modalities. To solve this issue, we introduce MoExtend, an effective framework designed to streamline the modality adaptation and extension of Mixture-of-Experts (MoE) models. MoExtend seamlessly integrates new experts into pre-trained MoE models, endowing them with novel knowledge without the need to tune pretrained models such as MoE and vision encoders. This approach enables rapid adaptation and extension to new modal data or tasks, effectively addressing the challenge of accommodating new modalities within LLMs. Furthermore, MoExtend avoids tuning pretrained models, thus mitigating the risk of catastrophic forgetting. Experimental results demonstrate the efficacy and efficiency of MoExtend in enhancing the multimodal capabilities of LLMs, contributing to advancements in multimodal AI research. | [
"Zhong, Shanshan",
"Gao, Shanghua",
"Huang, Zhongzhan",
"Wen, Wushao",
"Zitnik, Marinka",
"Zhou, Pan"
]
| {M}o{E}xtend: Tuning New Experts for Modality and Task Extension | acl-srw.8 | Poster | 2305.14839v2 |
https://aclanthology.org/2024.acl-srw.9.bib | @inproceedings{wang-chen-2024-interpretability,
title = "On the Interpretability of Deep Learning Models for Collaborative Argumentation Analysis in Classrooms",
author = "Wang, Deliang and
Chen, Gaowei",
editor = "Fu, Xiyan and
Fleisig, Eve",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 4: Student Research Workshop)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-srw.9",
pages = "92--102",
abstract = "Collaborative argumentation holds significant potential for enhancing students{'} learning outcomes within classroom settings. Consequently, researchers have explored the application of artificial intelligence (AI) to automatically analyze argumentation in these contexts. Despite the remarkable performance of deep learning models in this task, their lack of interpretability poses a critical challenge, leading to teachers{'} skepticism and limited utilization. To cultivate trust among teachers, this PhD thesis proposal aims to leverage explainable AI techniques to provide explanations for these deep learning models. Specifically, the study develops two deep learning models for automated analysis of argument moves (claim, evidence, and warrant) and specificity levels (low, medium, and high) within collaborative argumentation. To address the interpretability issue, four explainable AI methods are proposed: gradient sensitivity, gradient input, integrated gradient, and LIME. Computational experiments demonstrate the efficacy of these methods in elucidating model predictions by computing word contributions, with LIME delivering exceptional performance. Moreover, a quasi-experiment is designed to evaluate the impact of model explanations on user trust and knowledge, serving as a future study of this PhD proposal. By tackling the challenges of interpretability and trust, this PhD thesis proposal aims to contribute to fostering user trust in AI and facilitating the practical implementation of AI in educational contexts.",
}
| Collaborative argumentation holds significant potential for enhancing students{'} learning outcomes within classroom settings. Consequently, researchers have explored the application of artificial intelligence (AI) to automatically analyze argumentation in these contexts. Despite the remarkable performance of deep learning models in this task, their lack of interpretability poses a critical challenge, leading to teachers{'} skepticism and limited utilization. To cultivate trust among teachers, this PhD thesis proposal aims to leverage explainable AI techniques to provide explanations for these deep learning models. Specifically, the study develops two deep learning models for automated analysis of argument moves (claim, evidence, and warrant) and specificity levels (low, medium, and high) within collaborative argumentation. To address the interpretability issue, four explainable AI methods are proposed: gradient sensitivity, gradient input, integrated gradient, and LIME. Computational experiments demonstrate the efficacy of these methods in elucidating model predictions by computing word contributions, with LIME delivering exceptional performance. Moreover, a quasi-experiment is designed to evaluate the impact of model explanations on user trust and knowledge, serving as a future study of this PhD proposal. By tackling the challenges of interpretability and trust, this PhD thesis proposal aims to contribute to fostering user trust in AI and facilitating the practical implementation of AI in educational contexts. | [
"Wang, Deliang",
"Chen, Gaowei"
]
| On the Interpretability of Deep Learning Models for Collaborative Argumentation Analysis in Classrooms | acl-srw.9 | Poster | 2102.10293v1 |
https://aclanthology.org/2024.acl-srw.10.bib | @inproceedings{wang-etal-2024-document,
title = "Document Alignment based on Overlapping Fixed-Length Segments",
author = "Wang, Xiaotian and
Utsuro, Takehito and
Nagata, Masaaki",
editor = "Fu, Xiyan and
Fleisig, Eve",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 4: Student Research Workshop)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-srw.10",
pages = "103--113",
abstract = "Acquiring large-scale parallel corpora is crucial for NLP tasks such asNeural Machine Translation, and web crawling has become a popularmethodology for this purpose. Previous studies have been conductedbased on sentence-based segmentation (SBS) when aligning documents invarious languages which are obtained through web crawling. Among them,the TK-PERT method (Thompson and Koehn, 2020) achieved state-of-the-artresults and addressed the boilerplate text in web crawling data wellthrough a down-weighting approach. However, there remains a problemwith how to handle long-text encoding better. Thus, we introduce thestrategy of Overlapping Fixed-Length Segmentation (OFLS) in place ofSBS, and observe a pronounced enhancement when performing the sameapproach for document alignment. In this paper, we compare the SBS andOFLS using three previous methods, Mean-Pool, TK-PERT (Thompson andKoehn, 2020), and Optimal Transport (Clark et al., 2019; El- Kishky andGuzman, 2020), on the WMT16 document alignment shared task forFrench-English, as well as on our self-established Japanese-Englishdataset MnRN. As a result, for the WMT16 task, various SBS basedmethods showed an increase in recall by 1{\%} to 10{\%} after reproductionwith OFLS. For MnRN data, OFLS demonstrated notable accuracyimprovements and exhibited faster document embedding speed.",
}
| Acquiring large-scale parallel corpora is crucial for NLP tasks such asNeural Machine Translation, and web crawling has become a popularmethodology for this purpose. Previous studies have been conductedbased on sentence-based segmentation (SBS) when aligning documents invarious languages which are obtained through web crawling. Among them,the TK-PERT method (Thompson and Koehn, 2020) achieved state-of-the-artresults and addressed the boilerplate text in web crawling data wellthrough a down-weighting approach. However, there remains a problemwith how to handle long-text encoding better. Thus, we introduce thestrategy of Overlapping Fixed-Length Segmentation (OFLS) in place ofSBS, and observe a pronounced enhancement when performing the sameapproach for document alignment. In this paper, we compare the SBS andOFLS using three previous methods, Mean-Pool, TK-PERT (Thompson andKoehn, 2020), and Optimal Transport (Clark et al., 2019; El- Kishky andGuzman, 2020), on the WMT16 document alignment shared task forFrench-English, as well as on our self-established Japanese-Englishdataset MnRN. As a result, for the WMT16 task, various SBS basedmethods showed an increase in recall by 1{\%} to 10{\%} after reproductionwith OFLS. For MnRN data, OFLS demonstrated notable accuracyimprovements and exhibited faster document embedding speed. | [
"Wang, Xiaotian",
"Utsuro, Takehito",
"Nagata, Masaaki"
]
| Document Alignment based on Overlapping Fixed-Length Segments | acl-srw.10 | Poster | 2104.08777v1 |
https://aclanthology.org/2024.acl-srw.11.bib | @inproceedings{benedetti-etal-2024-automatically,
title = "Automatically Suggesting Diverse Example Sentences for {L}2 {J}apanese Learners Using Pre-Trained Language Models",
author = "Benedetti, Enrico and
Aizawa, Akiko and
Boudin, Florian",
editor = "Fu, Xiyan and
Fleisig, Eve",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 4: Student Research Workshop)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-srw.11",
pages = "114--131",
abstract = "Providing example sentences that are diverse and aligned with learners{'} proficiency levels is essential for fostering effective language acquisition.This study examines the use of Pre-trained Language Models (PLMs) to produce example sentences targeting L2 Japanese learners.We utilize PLMs in two ways: as quality scoring components in a retrieval system that draws from a newly curated corpus of Japanese sentences, and as direct sentence generators using zero-shot learning.We evaluate the quality of sentences by considering multiple aspects such as difficulty, diversity, and naturalness, with a panel of raters consisting of learners of Japanese, native speakers {--} and GPT-4.Our findings suggest that there is inherent disagreement among participants on the ratings of sentence qualities, except for difficulty. Despite that, the retrieval approach was preferred by all evaluators, especially for beginner and advanced target proficiency, while the generative approaches received lower scores on average.Even so, our experiments highlight the potential for using PLMs to enhance the adaptability of sentence suggestion systems and therefore improve the language learning journey.",
}
| Providing example sentences that are diverse and aligned with learners{'} proficiency levels is essential for fostering effective language acquisition.This study examines the use of Pre-trained Language Models (PLMs) to produce example sentences targeting L2 Japanese learners.We utilize PLMs in two ways: as quality scoring components in a retrieval system that draws from a newly curated corpus of Japanese sentences, and as direct sentence generators using zero-shot learning.We evaluate the quality of sentences by considering multiple aspects such as difficulty, diversity, and naturalness, with a panel of raters consisting of learners of Japanese, native speakers {--} and GPT-4.Our findings suggest that there is inherent disagreement among participants on the ratings of sentence qualities, except for difficulty. Despite that, the retrieval approach was preferred by all evaluators, especially for beginner and advanced target proficiency, while the generative approaches received lower scores on average.Even so, our experiments highlight the potential for using PLMs to enhance the adaptability of sentence suggestion systems and therefore improve the language learning journey. | [
"Benedetti, Enrico",
"Aizawa, Akiko",
"Boudin, Florian"
]
| Automatically Suggesting Diverse Example Sentences for {L}2 {J}apanese Learners Using Pre-Trained Language Models | acl-srw.11 | Poster | 2406.19650v1 |
https://aclanthology.org/2024.acl-srw.12.bib | @inproceedings{suwannapichat-etal-2024-z,
title = "{Z}-coref: {T}hai Coreference and Zero Pronoun Resolution",
author = "Suwannapichat, Poomphob and
Tarnpradab, Sansiri and
Prom-on, Santitham",
editor = "Fu, Xiyan and
Fleisig, Eve",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 4: Student Research Workshop)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-srw.12",
pages = "132--139",
abstract = "Coreference Resolution (CR) and Zero Pronoun Resolution (ZPR) are vital for extracting meaningful information from text. However, limited research and datasets pose significant challenges in Thai language. To address this, we developed an annotated joint CR and ZPR dataset. Additionally, we introduced the Z-coref model, capable of simultaneously handling CR and ZPR tasks by adjusting the span definition of a prior CR architecture to include token gaps. The proposed model trained on our dataset outperformed the state-of-the-art in resolving both coreference resolution and zero-pronoun resolution, while taking less time to train.",
}
| Coreference Resolution (CR) and Zero Pronoun Resolution (ZPR) are vital for extracting meaningful information from text. However, limited research and datasets pose significant challenges in Thai language. To address this, we developed an annotated joint CR and ZPR dataset. Additionally, we introduced the Z-coref model, capable of simultaneously handling CR and ZPR tasks by adjusting the span definition of a prior CR architecture to include token gaps. The proposed model trained on our dataset outperformed the state-of-the-art in resolving both coreference resolution and zero-pronoun resolution, while taking less time to train. | [
"Suwannapichat, Poomphob",
"Tarnpradab, Sansiri",
"Prom-on, Santitham"
]
| {Z}-coref: {T}hai Coreference and Zero Pronoun Resolution | acl-srw.12 | Poster | 9912004v1 |
https://aclanthology.org/2024.acl-srw.13.bib | @inproceedings{murali-etal-2024-remag,
title = "{R}e{MAG}-{KR}: Retrieval and Medically Assisted Generation with Knowledge Reduction for Medical Question Answering",
author = "Murali, Sidhaarth and
S., Sowmya and
R, Supreetha",
editor = "Fu, Xiyan and
Fleisig, Eve",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 4: Student Research Workshop)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-srw.13",
pages = "140--145",
abstract = "Large Language Models (LLMs) have significant potential for facilitating intelligent end-user applications in healthcare. However, hallucinations remain an inherent problem with LLMs, making it crucial to address this issue with extensive medical knowledge and data. In this work, we propose a Retrieve-and-Medically-Augmented-Generation with Knowledge Reduction (ReMAG-KR) pipeline, employing a carefully curated knowledge base using cross-encoder re-ranking strategies. The pipeline is tested on medical MCQ-based QA datasets as well as general QA datasets. It was observed that when the knowledge base is reduced, the model{'}s performance decreases by 2-8{\%}, while the inference time improves by 47{\%}.",
}
| Large Language Models (LLMs) have significant potential for facilitating intelligent end-user applications in healthcare. However, hallucinations remain an inherent problem with LLMs, making it crucial to address this issue with extensive medical knowledge and data. In this work, we propose a Retrieve-and-Medically-Augmented-Generation with Knowledge Reduction (ReMAG-KR) pipeline, employing a carefully curated knowledge base using cross-encoder re-ranking strategies. The pipeline is tested on medical MCQ-based QA datasets as well as general QA datasets. It was observed that when the knowledge base is reduced, the model{'}s performance decreases by 2-8{\%}, while the inference time improves by 47{\%}. | [
"Murali, Sidhaarth",
"S., Sowmya",
"R, Supreetha"
]
| {R}e{MAG}-{KR}: Retrieval and Medically Assisted Generation with Knowledge Reduction for Medical Question Answering | acl-srw.13 | Poster | 2310.01299v1 |
https://aclanthology.org/2024.acl-srw.14.bib | @inproceedings{xu-etal-2024-plot,
title = "Plot Retrieval as an Assessment of Abstract Semantic Association",
author = "Xu, Shicheng and
Pang, Liang and
Li, Jiangnan and
Yu, Mo and
Meng, Fandong and
Shen, Huawei and
Cheng, Xueqi and
Zhou, Jie",
editor = "Fu, Xiyan and
Fleisig, Eve",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 4: Student Research Workshop)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-srw.14",
pages = "146--161",
abstract = "Retrieving relevant plots from the book for a query is a critical task, which can improve the reading experience and efficiency of readers. Readers usually only give an abstract and vague description as the query based on their own understanding, summaries, or speculations of the plot, which requires the retrieval model to have a strong ability to estimate the abstract semantic associations between the query and candidate plots. However, existing information retrieval (IR) datasets cannot reflect this ability well. In this paper, we propose PlotRetrieval, a labeled dataset to train and evaluate the performance of IR models on the novel task Plot Retrieval. Text pairs in PlotRetrieval have less word overlap and more abstract semantic association, which can reflect the ability of the IR models to estimate the abstract semantic association, rather than just traditional lexical or semantic matching. Extensive experiments across various lexical retrieval, sparse retrieval, dense retrieval, and cross-encoder methods compared with human studies on PlotRetrieval show current IR models still struggle in capturing abstract semantic association between texts. PlotRetrieval can be the benchmark for further research on the semantic association modeling ability of IR models.",
}
| Retrieving relevant plots from the book for a query is a critical task, which can improve the reading experience and efficiency of readers. Readers usually only give an abstract and vague description as the query based on their own understanding, summaries, or speculations of the plot, which requires the retrieval model to have a strong ability to estimate the abstract semantic associations between the query and candidate plots. However, existing information retrieval (IR) datasets cannot reflect this ability well. In this paper, we propose PlotRetrieval, a labeled dataset to train and evaluate the performance of IR models on the novel task Plot Retrieval. Text pairs in PlotRetrieval have less word overlap and more abstract semantic association, which can reflect the ability of the IR models to estimate the abstract semantic association, rather than just traditional lexical or semantic matching. Extensive experiments across various lexical retrieval, sparse retrieval, dense retrieval, and cross-encoder methods compared with human studies on PlotRetrieval show current IR models still struggle in capturing abstract semantic association between texts. PlotRetrieval can be the benchmark for further research on the semantic association modeling ability of IR models. | [
"Xu, Shicheng",
"Pang, Liang",
"Li, Jiangnan",
"Yu, Mo",
"Meng, F",
"ong",
"Shen, Huawei",
"Cheng, Xueqi",
"Zhou, Jie"
]
| Plot Retrieval as an Assessment of Abstract Semantic Association | acl-srw.14 | Poster | 2311.01666v1 |
https://aclanthology.org/2024.acl-srw.15.bib | @inproceedings{wang-etal-2024-demystifying,
title = "Demystifying Instruction Mixing for Fine-tuning Large Language Models",
author = "Wang, Renxi and
Li, Haonan and
Wu, Minghao and
Wang, Yuxia and
Han, Xudong and
Zhang, Chiyu and
Baldwin, Timothy",
editor = "Fu, Xiyan and
Fleisig, Eve",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 4: Student Research Workshop)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-srw.15",
pages = "162--169",
abstract = "Instruction tuning significantly enhances the performance of large language models (LLMs) across various tasks. However, the procedure to optimizing the mixing of instruction datasets for LLM fine-tuning is still poorly understood. This study categorizes instructions into three primary types: NLP downstream tasks, coding, and general chat. We explore the effects of instruction tuning on different combinations of datasets on LLM performance, and find that certain instruction types are more advantageous for specific applications but can negatively impact other areas. This work provides insights into instruction mixtures, laying the foundations for future research.",
}
| Instruction tuning significantly enhances the performance of large language models (LLMs) across various tasks. However, the procedure to optimizing the mixing of instruction datasets for LLM fine-tuning is still poorly understood. This study categorizes instructions into three primary types: NLP downstream tasks, coding, and general chat. We explore the effects of instruction tuning on different combinations of datasets on LLM performance, and find that certain instruction types are more advantageous for specific applications but can negatively impact other areas. This work provides insights into instruction mixtures, laying the foundations for future research. | [
"Wang, Renxi",
"Li, Haonan",
"Wu, Minghao",
"Wang, Yuxia",
"Han, Xudong",
"Zhang, Chiyu",
"Baldwin, Timothy"
]
| Demystifying Instruction Mixing for Fine-tuning Large Language Models | acl-srw.15 | Poster | 2312.10793v3 |
https://aclanthology.org/2024.acl-srw.16.bib | @inproceedings{mainzinger-levow-2024-fine,
title = "Fine-Tuning {ASR} models for Very Low-Resource Languages: A Study on Mvskoke",
author = "Mainzinger, Julia and
Levow, Gina-Anne",
editor = "Fu, Xiyan and
Fleisig, Eve",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 4: Student Research Workshop)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-srw.16",
pages = "170--176",
abstract = "Recent advancements in multilingual models for automatic speech recognition (ASR) have been able to achieve a high accuracy for languages with extremely limited resources. This study examines ASR modeling for the Mvskoke language, an indigenous language of America. The parameter efficiency of adapter training is contrasted with training entire models, and it is demonstrated how performance varies with different amounts of data. Additionally, the models are evaluated with trigram language model decoding, and the outputs are compared across different types of speech recordings. Results show that training an adapter is both parameter efficient and gives higher accuracy for a relatively small amount of data.",
}
| Recent advancements in multilingual models for automatic speech recognition (ASR) have been able to achieve a high accuracy for languages with extremely limited resources. This study examines ASR modeling for the Mvskoke language, an indigenous language of America. The parameter efficiency of adapter training is contrasted with training entire models, and it is demonstrated how performance varies with different amounts of data. Additionally, the models are evaluated with trigram language model decoding, and the outputs are compared across different types of speech recordings. Results show that training an adapter is both parameter efficient and gives higher accuracy for a relatively small amount of data. | [
"Mainzinger, Julia",
"Levow, Gina-Anne"
]
| Fine-Tuning {ASR} models for Very Low-Resource Languages: A Study on Mvskoke | acl-srw.16 | Poster | 1809.01431v2 |
https://aclanthology.org/2024.acl-srw.17.bib | @inproceedings{parfenova-etal-2024-automating,
title = "Automating Qualitative Data Analysis with Large Language Models",
author = {Parfenova, Angelina and
Denzler, Alexander and
Pfeffer, J{\"o}rgen},
editor = "Fu, Xiyan and
Fleisig, Eve",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 4: Student Research Workshop)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-srw.17",
pages = "177--185",
abstract = "This PhD proposal aims to investigate ways of automating qualitative data analysis, specifically the thematic coding of texts. Despite existing methods vastly covered in literature, they mainly use Topic Modeling and other quantitative approaches which are far from resembling a human{'}s analysis outcome. This proposal examines the limitations of current research in the field. It proposes a novel methodology based on Large Language Models to tackle automated coding and make it as close as possible to the results of human researchers. This paper covers studies already done in this field and their limitations, existing software, the problem of duplicating the researcher bias, and the proposed methodology.",
}
| This PhD proposal aims to investigate ways of automating qualitative data analysis, specifically the thematic coding of texts. Despite existing methods vastly covered in literature, they mainly use Topic Modeling and other quantitative approaches which are far from resembling a human{'}s analysis outcome. This proposal examines the limitations of current research in the field. It proposes a novel methodology based on Large Language Models to tackle automated coding and make it as close as possible to the results of human researchers. This paper covers studies already done in this field and their limitations, existing software, the problem of duplicating the researcher bias, and the proposed methodology. | [
"Parfenova, Angelina",
"Denzler, Alex",
"er",
"Pfeffer, J{\\\"o}rgen"
]
| Automating Qualitative Data Analysis with Large Language Models | acl-srw.17 | Poster | 2403.08844v1 |
https://aclanthology.org/2024.acl-srw.18.bib | @inproceedings{herrlein-etal-2024-anhalten,
title = "{ANHALTEN}: Cross-Lingual Transfer for {G}erman Token-Level Reference-Free Hallucination Detection",
author = "Herrlein, Janek and
Hung, Chia-Chien and
Glava�, Goran",
editor = "Fu, Xiyan and
Fleisig, Eve",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 4: Student Research Workshop)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-srw.18",
pages = "186--194",
abstract = "Research on token-level reference-free hallucination detection has predominantly focused on English, primarily due to the scarcity of robust datasets in other languages. This has hindered systematic investigations into the effectiveness of cross-lingual transfer for this important NLP application. To address this gap, we introduce ANHALTEN, a new evaluation dataset that extends the English hallucination detection dataset to German. To the best of our knowledge, this is the first work that explores cross-lingual transfer for token-level reference-free hallucination detection. ANHALTEN contains gold annotations in German that are parallel (i.e., directly comparable to the original English instances). We benchmark several prominent cross-lingual transfer approaches, demonstrating that larger context length leads to better hallucination detection in German, even without succeeding context. Importantly, we show that the sample-efficient few-shot transfer is the most effective approach in most setups. This highlights the practical benefits of minimal annotation effort in the target language for reference-free hallucination detection. Aiming to catalyze future research on cross-lingual token-level reference-free hallucination detection, we make ANHALTEN publicly available: https://github.com/janekh24/anhalten",
}
| Research on token-level reference-free hallucination detection has predominantly focused on English, primarily due to the scarcity of robust datasets in other languages. This has hindered systematic investigations into the effectiveness of cross-lingual transfer for this important NLP application. To address this gap, we introduce ANHALTEN, a new evaluation dataset that extends the English hallucination detection dataset to German. To the best of our knowledge, this is the first work that explores cross-lingual transfer for token-level reference-free hallucination detection. ANHALTEN contains gold annotations in German that are parallel (i.e., directly comparable to the original English instances). We benchmark several prominent cross-lingual transfer approaches, demonstrating that larger context length leads to better hallucination detection in German, even without succeeding context. Importantly, we show that the sample-efficient few-shot transfer is the most effective approach in most setups. This highlights the practical benefits of minimal annotation effort in the target language for reference-free hallucination detection. Aiming to catalyze future research on cross-lingual token-level reference-free hallucination detection, we make ANHALTEN publicly available: https://github.com/janekh24/anhalten | [
"Herrlein, Janek",
"Hung, Chia-Chien",
"Glava�, Goran"
]
| {ANHALTEN}: Cross-Lingual Transfer for {G}erman Token-Level Reference-Free Hallucination Detection | acl-srw.18 | Poster | 2407.13702v1 |
https://aclanthology.org/2024.acl-srw.19.bib | @inproceedings{thaminkaew-etal-2024-label,
title = "Label-Aware Automatic Verbalizer for Few-Shot Text Classification in Mid-To-Low Resource Languages",
author = "Thaminkaew, Thanakorn and
Lertvittayakumjorn, Piyawat and
Vateekul, Peerapon",
editor = "Fu, Xiyan and
Fleisig, Eve",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 4: Student Research Workshop)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-srw.19",
pages = "195--203",
abstract = "Prompt-based learning has shown its effectiveness in few-shot text classification. A key factor in its success is a verbalizer, which translates output from a language model into a predicted class. Notably, the simplest and widely acknowledged verbalizer employs manual labels to represent the classes. However, manual selection may not yield the optimal words for a given language model, potentially leading to subpar classification performance, especially in mid-to-low resource languages with weaker language models. Therefore, we propose Label-Aware Automatic Verbalizer (LAAV), effectively augmenting manual labels for improved few-shot classification results. Specifically, we utilize the label name along with the conjunction {``}and{''} to induce the model to generate more effective words for the verbalizer. Experimental results on four mid-to-low resource Southeast Asian languages demonstrate that LAAV significantly outperforms existing verbalizers.",
}
| Prompt-based learning has shown its effectiveness in few-shot text classification. A key factor in its success is a verbalizer, which translates output from a language model into a predicted class. Notably, the simplest and widely acknowledged verbalizer employs manual labels to represent the classes. However, manual selection may not yield the optimal words for a given language model, potentially leading to subpar classification performance, especially in mid-to-low resource languages with weaker language models. Therefore, we propose Label-Aware Automatic Verbalizer (LAAV), effectively augmenting manual labels for improved few-shot classification results. Specifically, we utilize the label name along with the conjunction {``}and{''} to induce the model to generate more effective words for the verbalizer. Experimental results on four mid-to-low resource Southeast Asian languages demonstrate that LAAV significantly outperforms existing verbalizers. | [
"Thaminkaew, Thanakorn",
"Lertvittayakumjorn, Piyawat",
"Vateekul, Peerapon"
]
| Label-Aware Automatic Verbalizer for Few-Shot Text Classification in Mid-To-Low Resource Languages | acl-srw.19 | Poster | 2310.12778v1 |
https://aclanthology.org/2024.acl-srw.20.bib | @inproceedings{estve-etal-2024-vector,
title = "Vector Spaces for Quantifying Disparity of Multiword Expressions in Annotated Text",
author = "Est�ve, Louis and
Savary, Agata and
Lavergne, Thomas",
editor = "Fu, Xiyan and
Fleisig, Eve",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 4: Student Research Workshop)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-srw.20",
pages = "204--224",
abstract = "Multiword Expressions (MWEs) make a goodcase study for linguistic diversity due to theiridiosyncratic nature. Defining MWE canonicalforms as types, diversity may be measurednotably through disparity, based on pairwisedistances between types. To this aim, wetrain static MWE-aware word embeddings forverbal MWEs in 14 languages, and we showinteresting properties of these vector spaces.We use these vector spaces to implement theso-called functional diversity measure. Weapply this measure to the results of severalMWE identification systems. We find that,although MWE vector spaces are meaningful ata local scale, the disparity measure aggregatingthem at a global scale strongly correlateswith the number of types, which questions itsusefulness in presence of simpler diversitymetrics such as variety. We make the vectorspaces we generated available.",
}
| Multiword Expressions (MWEs) make a goodcase study for linguistic diversity due to theiridiosyncratic nature. Defining MWE canonicalforms as types, diversity may be measurednotably through disparity, based on pairwisedistances between types. To this aim, wetrain static MWE-aware word embeddings forverbal MWEs in 14 languages, and we showinteresting properties of these vector spaces.We use these vector spaces to implement theso-called functional diversity measure. Weapply this measure to the results of severalMWE identification systems. We find that,although MWE vector spaces are meaningful ata local scale, the disparity measure aggregatingthem at a global scale strongly correlateswith the number of types, which questions itsusefulness in presence of simpler diversitymetrics such as variety. We make the vectorspaces we generated available. | [
"Est�ve, Louis",
"Savary, Agata",
"Lavergne, Thomas"
]
| Vector Spaces for Quantifying Disparity of Multiword Expressions in Annotated Text | acl-srw.20 | Poster | 1612.00246v1 |
https://aclanthology.org/2024.acl-srw.21.bib | @inproceedings{sinelnik-hovy-2024-narratives,
title = "Narratives at Conflict: Computational Analysis of News Framing in Multilingual Disinformation Campaigns",
author = "Sinelnik, Antonina and
Hovy, Dirk",
editor = "Fu, Xiyan and
Fleisig, Eve",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 4: Student Research Workshop)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-srw.21",
pages = "225--237",
abstract = "Any report frames issues to favor a particular interpretation by highlighting or excluding certain aspects of a story. Despite the widespread use of framing in disinformation, framing properties and detection methods remain underexplored outside the English-speaking world. We explore how multilingual framing of the same issue differs systematically. We use eight years of Russia-backed disinformation campaigns, spanning 8k news articles in 4 languages targeting 15 countries. We find that disinformation campaigns consistently and intentionally favor specific framing, depending on the target language of the audience. We further discover how Russian-language articles consistently highlight selected frames depending on the region of the media coverage. We find that the two most prominent models for automatic frame analysis underperform and show high disagreement, highlighting the need for further research.",
}
| Any report frames issues to favor a particular interpretation by highlighting or excluding certain aspects of a story. Despite the widespread use of framing in disinformation, framing properties and detection methods remain underexplored outside the English-speaking world. We explore how multilingual framing of the same issue differs systematically. We use eight years of Russia-backed disinformation campaigns, spanning 8k news articles in 4 languages targeting 15 countries. We find that disinformation campaigns consistently and intentionally favor specific framing, depending on the target language of the audience. We further discover how Russian-language articles consistently highlight selected frames depending on the region of the media coverage. We find that the two most prominent models for automatic frame analysis underperform and show high disagreement, highlighting the need for further research. | [
"Sinelnik, Antonina",
"Hovy, Dirk"
]
| Narratives at Conflict: Computational Analysis of News Framing in Multilingual Disinformation Campaigns | acl-srw.21 | Poster | 2406.15443v1 |
https://aclanthology.org/2024.acl-srw.22.bib | @inproceedings{schelb-etal-2024-assessing,
title = "Assessing In-context Learning and Fine-tuning for Topic Classification of {G}erman Web Data",
author = "Schelb, Julian and
Spitz, Andreas and
Ulloa, Roberto",
editor = "Fu, Xiyan and
Fleisig, Eve",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 4: Student Research Workshop)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-srw.22",
pages = "238--252",
abstract = "Researchers in the political and social sciences often rely on classification models to analyze trends in information consumption by examining browsing histories of millions of webpages. Automated scalable methods are necessary due to the impracticality of manual labeling. In this paper, we model the detection of topic-related content as a binary classification task and compare the accuracy of fine-tuned pre-trained encoder models against in-context learning strategies. Using only a few hundred annotated data points per topic, we detect content related to three German policies in a database of scraped webpages. We compare multilingual and monolingual models, as well as zero and few-shot approaches, and investigate the impact of negative sampling strategies and the combination of URL {\&} content-based features. Our results show that a small sample of annotated data is sufficient to train an effective classifier. Fine-tuning encoder-based models yields better results than in-context learning. Classifiers using both URL {\&} content-based features perform best, while using URLs alone provides adequate results when content is unavailable.",
}
| Researchers in the political and social sciences often rely on classification models to analyze trends in information consumption by examining browsing histories of millions of webpages. Automated scalable methods are necessary due to the impracticality of manual labeling. In this paper, we model the detection of topic-related content as a binary classification task and compare the accuracy of fine-tuned pre-trained encoder models against in-context learning strategies. Using only a few hundred annotated data points per topic, we detect content related to three German policies in a database of scraped webpages. We compare multilingual and monolingual models, as well as zero and few-shot approaches, and investigate the impact of negative sampling strategies and the combination of URL {\&} content-based features. Our results show that a small sample of annotated data is sufficient to train an effective classifier. Fine-tuning encoder-based models yields better results than in-context learning. Classifiers using both URL {\&} content-based features perform best, while using URLs alone provides adequate results when content is unavailable. | [
"Schelb, Julian",
"Spitz, Andreas",
"Ulloa, Roberto"
]
| Assessing In-context Learning and Fine-tuning for Topic Classification of {G}erman Web Data | acl-srw.22 | Poster | 2407.16516v1 |
https://aclanthology.org/2024.acl-srw.23.bib | @inproceedings{ishigaki-etal-2024-knowledge,
title = "Knowledge Editing of Large Language Models Unconstrained by Word Order",
author = "Ishigaki, Ryoma and
Suzuki, Jundai and
Shuzo, Masaki and
Maeda, Eisaku",
editor = "Fu, Xiyan and
Fleisig, Eve",
booktitle = "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 4: Student Research Workshop)",
month = aug,
year = "2024",
address = "Bangkok, Thailand",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.acl-srw.23",
pages = "253--263",
abstract = "Large Language Models (LLMs) are considered to have potentially extensive knowledge, but because their internal processing is black-boxed, it has been difficult to directly edit the knowledge held by the LLMs themselves. To address this issue, a method called local modification-based knowledge editing has been developed. This method identifies the knowledge neurons that encode the target knowledge and adjusts the parameters associated with these neurons to update the knowledge. Knowledge neurons are identified by masking the $\it{o}$ part from sentences representing relational triplets ($\it{s, r, o}$), having the LLM predict the masked part, and observing the LLM�s activation during the prediction. When the architecture is decoder-based, the predicted $\it{o}$ needs to be located at the end of the sentence. Previous local modification-based knowledge editing methods for decoder-based models have assumed SVO languages and faced challenges when applied to SOV languages such as Japanese. In this study, we propose a knowledge editing method that eliminates the need for word order constraints by converting the input for identifying knowledge neurons into a question where $\it{o}$ is the answer. We conducted validation experiments on 500 examples and confirmed that the proposed method is effective for Japanese, a non-SVO language. We also applied this method to English, an SVO language, and demonstrated that it outperforms conventional methods.",
}
| Large Language Models (LLMs) are considered to have potentially extensive knowledge, but because their internal processing is black-boxed, it has been difficult to directly edit the knowledge held by the LLMs themselves. To address this issue, a method called local modification-based knowledge editing has been developed. This method identifies the knowledge neurons that encode the target knowledge and adjusts the parameters associated with these neurons to update the knowledge. Knowledge neurons are identified by masking the $\it{o}$ part from sentences representing relational triplets ($\it{s, r, o}$), having the LLM predict the masked part, and observing the LLM�s activation during the prediction. When the architecture is decoder-based, the predicted $\it{o}$ needs to be located at the end of the sentence. Previous local modification-based knowledge editing methods for decoder-based models have assumed SVO languages and faced challenges when applied to SOV languages such as Japanese. In this study, we propose a knowledge editing method that eliminates the need for word order constraints by converting the input for identifying knowledge neurons into a question where $\it{o}$ is the answer. We conducted validation experiments on 500 examples and confirmed that the proposed method is effective for Japanese, a non-SVO language. We also applied this method to English, an SVO language, and demonstrated that it outperforms conventional methods. | [
"Ishigaki, Ryoma",
"Suzuki, Jundai",
"Shuzo, Masaki",
"Maeda, Eisaku"
]
| Knowledge Editing of Large Language Models Unconstrained by Word Order | acl-srw.23 | Poster | 2402.13593v1 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.