Routing Drift Alone Does Not Diagnose Failure in Merged MoE LLMs Paper • 2609.32821 • Published 11 days ago • 5
Not All Disagreement Is Learnable: Token Teachability in On-Policy Distillation Paper • 2605.26844 • Published May 26 • 24
Geometry Conflict: Explaining and Controlling Forgetting in LLM Continual Post-Training Paper • 2605.09608 • Published May 10 • 52