OldFriends (ZHANG Jipeng)

upvoted a collection 10 days ago

MIT Talk 31/10 Papers

Collection

14 items • Updated 13 days ago • 26

upvoted a collection 19 days ago

LLaVA-Critic

Collection

as a general evaluator for assessing model performance • 6 items • Updated Oct 6 • 7

upvoted a paper about 1 month ago

Personalized Visual Instruction Tuning

Paper • 2410.07113 • Published Oct 9 • 69

upvoted an article 3 months ago

Article

SmolLM - blazingly fast and remarkably powerful

Jul 16

• 258

upvoted a collection 4 months ago

NuminaMath

Collection

Datasets and models for training SOTA math LLMs. See our GitHub for training & inference code: https://github.com/project-numina/aimo-progress-prize • 6 items • Updated Jul 21 • 62

upvoted an article 4 months ago

Article

How NuminaMath Won the 1st AIMO Progress Prize

Jul 11

• 100

upvoted a paper 4 months ago

TheoremLlama: Transforming General-Purpose LLMs into Lean4 Experts

Paper • 2407.03203 • Published Jul 3 • 10

upvoted an article 4 months ago

Article

Large-scale Near-deduplication Behind BigCode

May 16, 2023

• 18

upvoted a paper 5 months ago

Jailbreaking as a Reward Misspecification Problem

Paper • 2406.14393 • Published Jun 20 • 12

upvoted a paper 6 months ago

RLHF Workflow: From Reward Modeling to Online RLHF

Paper • 2405.07863 • Published May 13 • 67

upvoted a paper 8 months ago

LISA: Layerwise Importance Sampling for Memory-Efficient Large Language Model Fine-Tuning

Paper • 2403.17919 • Published Mar 26 • 16

ZHANG Jipeng

AI & ML interests

Organizations

OldFriends's activity

MIT Talk 31/10 Papers

LLaVA-Critic

Personalized Visual Instruction Tuning

SmolLM - blazingly fast and remarkably powerful

NuminaMath

How NuminaMath Won the 1st AIMO Progress Prize

TheoremLlama: Transforming General-Purpose LLMs into Lean4 Experts

Large-scale Near-deduplication Behind BigCode

Jailbreaking as a Reward Misspecification Problem

RLHF Workflow: From Reward Modeling to Online RLHF

LISA: Layerwise Importance Sampling for Memory-Efficient Large Language Model Fine-Tuning