The Teacher Is a Direction, Not a Destination: Extrapolating RL-Induced Representation Residuals in On-Policy Distillation Paper • 2609.36484 • Published 8 days ago • 578
False Frontiers: Diagnosing and Mitigating Co-Cheating in Self-Evolving Search Agents Paper • 2609.39102 • Published 7 days ago • 671
Compile by Training: Turning Natural-Language Specifications into Local Neural Functions Paper • 2609.04199 • Published Sep 3 • 333