Shenzhi Yang

Shenzhi's activity

• All
• Models
• Datasets
• Spaces
• Buckets
• Papers
• Collections
• Community
• Posts
• Upvotes
• Likes
• Articles

upvoted a paper 2 days ago

The Teacher Is a Direction, Not a Destination: Extrapolating RL-Induced Representation Residuals in On-Policy Distillation

Paper • 2609.36484 • Published 26 days ago • 208

upvoted 3 papers 5 days ago

Vision-Language-Delta-Action: Learning Compact Deltas as Actionable Transition Interfaces for Robotic Manipulation

Paper • 2610.25874 • Published 6 days ago • 5

World Model Interpretability: From Decoding Representations to Understanding Latent Worlds

Paper • 2610.25296 • Published 7 days ago • 9

Text-to-Image Generative Modeling Without Language Pretraining

Paper • 2610.26863 • Published 6 days ago • 8

upvoted a paper 25 days ago

Look Before You Select: Rethinking Vocabulary Sparsification in On-Policy Distillation

Paper • 2609.34386 • Published 27 days ago • 8

upvoted a collection 5 months ago

WTF GENIUS PAPERS

Collection

Papers that made me appreciate my major and my life a little more. obs=Observation, innov=Innovation. Most papers are abt improving tiny models. • 489 items • Updated 4 days ago • 104

upvoted a paper 5 months ago

OPRD: On-Policy Representation Distillation

Paper • 2606.06021 • Published Jun 4 • 8

upvoted an article 6 months ago

view article

Article

From GRPO to DAPO and GSPO: What, Why, and How

• 

NormalUhr

•

Aug 9, 2025

• 148

upvoted a paper 7 months ago

Can LLMs Learn to Reason Robustly under Noisy Supervision?

Paper • 2604.03993 • Published Apr 5 • 39

upvoted a paper 9 months ago

Your Group-Relative Advantage Is Biased

Paper • 2601.08521 • Published Jan 13 • 158

upvoted a paper 10 months ago

TraPO: A Semi-Supervised Reinforcement Learning Framework for Boosting LLM Reasoning

Paper • 2512.13106 • Published Dec 15, 2025 • 4