---
license: mit
language:
- en
---
## Introduction

- Paper: [Empowering Reliable Visual-Centric Instruction Following in MLLMs](https://arxiv.org/abs/2601.03198)
- Github: [KerenWLHe/VC-IFEval](https://github.com/KerenWLHe/VC-IFEval)

😊 This is the official repo of **VC-IFEngine datasets**.

📊 Image sources: [LLaVA-Instruct](https://huggingface.co/datasets/liuhaotian/LLaVA-Instruct-150K), [Visual Genome](https://homes.cs.washington.edu/~ranjay/visualgenome/), [ALLaVA-4V](https://huggingface.co/datasets/FreedomIntelligence/ALLaVA-4V)

## Format


**SFT** — `messages` format (role: user / assistant):

```json
{
  "messages": [
    {"role": "user", "content": "<image>\n{instruction}"},
    {"role": "assistant", "content": "{response}"}
  ],
  "images": ["Data/CC3M/images/example.jpg"]
}
```

**DPO** — `conversations` + `chosen` / `rejected` format:

```json
{
  "conversations": [
    {"from": "human", "value": "<image>\n{instruction}"}
  ],
  "chosen": {"from": "gpt", "value": "{full-constraint response}"},
  "rejected": {"from": "gpt", "value": "{constraint-degraded response}"},
  "images": ["Data/CC3M/images/example.jpg"]
}
```


## Citation

```bibtex
@misc{he2026empoweringreliablevisualcentricinstruction,
      title={Empowering Reliable Visual-Centric Instruction Following in MLLMs}, 
      author={Weilei He and Feng Ju and Zhiyuan Fan and Rui Min and Minhao Cheng and Yi R. Fung},
      year={2026},
      eprint={2601.03198},
      archivePrefix={arXiv},
      primaryClass={cs.LG},
      url={https://arxiv.org/abs/2601.03198}, 
}
```