The Teacher Is a Direction, Not a Destination: Extrapolating RL-Induced Representation Residuals in On-Policy Distillation
Paper • 2609.36484 • Published • 511
Ha-eun Cho
shubhamdasko
·
AI & ML interests
Interested in machine learning, NLP and open source AI.
Recent Activity
liked a model 5 days ago
drbaph/MiniMax-H3-Turbo-Lora-ComfyUI upvoted a paper 5 days ago
Does Learning Protein Folding Generalize to Broader Reasoning?Organizations
None yet
upvoted 3
papers 5 days ago