DRIFT: Decoupled Rollouts and Importance-Weighted Fine-Tuning for Efficient Multi-Turn Optimization Paper • 2605.31455 • Published May 29 • 6
Crafter: A Multi-Agent Harness for Editable Scientific Figure Generation from Diverse Inputs Paper • 2605.30611 • Published May 28 • 250
Gamma-World: Generative Multi-Agent World Modeling Beyond Two Players Paper • 2605.28816 • Published May 27 • 431
ControlLight: Towards Controllable, Consistent, and Generalizable Low-Light Enhancement Paper • 2605.25569 • Published May 25 • 21
Anti-Self-Distillation for Reasoning RL via Pointwise Mutual Information Paper • 2605.11609 • Published May 12 • 196
Matérn Noise for Triangulation-Agnostic Flow Matching on Meshes Paper • 2605.19305 • Published May 19 • 6
CiteVQA: Benchmarking Evidence Attribution for Trustworthy Document Intelligence Paper • 2605.12882 • Published May 13 • 274
Learning to Foresee: Unveiling the Unlocking Efficiency of On-Policy Distillation Paper • 2605.11739 • Published May 13 • 60
EvolveMem:Self-Evolving Memory Architecture via AutoResearch for LLM Agents Paper • 2605.13941 • Published May 13 • 24
SEIF: Self-Evolving Reinforcement Learning for Instruction Following Paper • 2605.07465 • Published May 8 • 30
MiA-Signature: Approximating Global Activation for Long-Context Understanding Paper • 2605.06416 • Published May 7 • 57