The Teacher Is a Direction, Not a Destination: Extrapolating RL-Induced Representation Residuals in On-Policy Distillation
Paper • 2609.36484 • Published • 556
Megan Garcia
megangar
·
AI & ML interests
natural language understanding, reinforcement learning, mechanistic interpretability, question answering
Recent Activity
upvoted a paper 11 days ago
False Frontiers: Diagnosing and Mitigating Co-Cheating in Self-Evolving Search Agents updated a dataset 22 days ago
megangar/legal-collection-2023Organizations
None yet
upvoted 2
papers 11 days ago