GroundAnything: Reconciling Parallel Decoding with Precise Visual Grounding at Flash Speed Paper • 2609.39600 • Published Sep 30 • 34
False Frontiers: Diagnosing and Mitigating Co-Cheating in Self-Evolving Search Agents Paper • 2609.39102 • Published Sep 30 • 151
The Teacher Is a Direction, Not a Destination: Extrapolating RL-Induced Representation Residuals in On-Policy Distillation Paper • 2609.36484 • Published Sep 29 • 168
Unlocking Lossless Speedups in LLMs via Discrete Diffusion Paper • 2609.04010 • Published Sep 3 • 113