The Teacher Is a Direction, Not a Destination: Extrapolating RL-Induced Representation Residuals in On-Policy Distillation
Paper • 2609.36484 • Published • 168
Nicholas
Rutger-sbiolab
AI & ML interests
When I'm not coding, I'm probably cooking
Recent Activity
upvoted a paper about 1 month ago
False Frontiers: Diagnosing and Mitigating Co-Cheating in Self-Evolving Search Agents upvoted a paper about 1 month ago
Transformers Stop Thinking Too Early, and a Tiny LoRA Fixes ItOrganizations
None yet
upvoted 4
papers about 1 month ago
False Frontiers: Diagnosing and Mitigating Co-Cheating in Self-Evolving Search Agents
Paper • 2609.39102 • Published • 151
Transformers Stop Thinking Too Early, and a Tiny LoRA Fixes It
Paper • 2609.36585 • Published • 27
MaLiang-Harness: A Programmable Path to Image and Video Generation
Paper • 2609.34309 • Published • 89