False Frontiers: Diagnosing and Mitigating Co-Cheating in Self-Evolving Search Agents Paper • 2609.39102 • Published 7 days ago • 672
Transformers Stop Thinking Too Early, and a Tiny LoRA Fixes It Paper • 2609.36585 • Published 8 days ago • 76
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
FlowBalance: Verifier-Grounded Self-Improvement from On-Policy Reasoning Experience Paper • 2609.03241 • Published Sep 3 • 54