LongStraw: Long-Context RL Beyond 2M Tokens under a Fixed GPU Budget Paper • 2607.14952 • Published 9 days ago • 197
SEED: Self-Evolving On-Policy Distillation for Agentic Reinforcement Learning Paper • 2607.14777 • Published 9 days ago • 99
Scaling the Horizon, Not the Parameters: Reaching Trillion-Parameter Performance with a 35B Agent Paper • 2606.30616 • Published 26 days ago • 103
LoopCoder-v2: Only Loop Once for Efficient Test-Time Computation Scaling Paper • 2606.18023 • Published Jun 16 • 210
Dense Supervision, Sparse Updates: On the Sparsity and Geometry of On-Policy Distillation Paper • 2606.13657 • Published Jun 11 • 5
WeaveBench: A Long-Horizon, Real-World Benchmark for Computer-Use Agents with Hybrid Interfaces Paper • 2606.09426 • Published Jun 8 • 106
Gamma-World: Generative Multi-Agent World Modeling Beyond Two Players Paper • 2605.28816 • Published May 27 • 433