Keane Moraes's picture

1 8 19

Keane Moraes

lordvader31

·

https://itskeane.info

kj3moraes

AI & ML interests

Knowledge graphs, NLP and PKM

Recent Activity

liked a dataset about 2 months ago

HuggingFaceTB/smoltalk

liked a model 3 months ago

meta-llama/Llama-3.2-1B

reacted to mlabonne's post with 👍 6 months ago

Large models are surprisingly bad storytellers. I asked 8 LLMs to "Tell me a bedtime story about bears and waffles." Claude 3.5 Sonnet and GPT-4o gave me the worst stories: no conflict, no moral, zero creativity. In contrast, smaller models were quite creative and wrote stories involving talking waffle trees and bears ostracized for their love of waffles. Here you can see a comparison between Claude 3.5 Sonnet and NeuralDaredevil-8B-abliterated. They both start with a family of bears but quickly diverge in terms of personality, conflict, etc. I mapped it to the hero's journey to have some kind of framework. Prompt engineering can definitely help here, but it's still disappointing that the larger models don't create better stories right off the bat. Do you know why smaller models outperform the frontier models here?

View all activity

Organizations

None yet

lordvader31's activity

upvoted a paper 7 months ago

Let the Expert Stick to His Last: Expert-Specialized Fine-Tuning for Sparse Architectural Large Language Models

Paper • 2407.01906 • Published Jul 2, 2024 • 35

upvoted a collection 7 months ago

Code Llama Family

This collection hosts the transformers repos of the Code Llama release • 12 items • Updated Dec 6, 2024 • 42

upvoted an article 8 months ago

Article

Finetune Mixtral 8x7B with AutoTrain

By

•

Apr 1, 2024

• 8

upvoted a collection 9 months ago

🔍 Interpretability & Analysis of LMs

Outstanding research in LM interpretability and evaluation, summarized • 95 items • Updated 3 days ago • 96

upvoted a collection 10 months ago

Image-to-Text Models 📝

This collection contains image captioning and OCR models. • 15 items • Updated Sep 19, 2023 • 7

upvoted 2 papers 10 months ago

Enhancing Vision-Language Pre-training with Rich Supervisions

Paper • 2403.03346 • Published Mar 5, 2024 • 16

Chatbot Arena: An Open Platform for Evaluating LLMs by Human Preference

Paper • 2403.04132 • Published Mar 7, 2024 • 38

upvoted a collection 12 months ago

MoEs papers reading list

60 items • Updated Nov 4, 2024 • 137