StateM: Reaching 95.3% Raw Accuracy, or a \$15 Frontier Run, on Terminal-Bench 2.1 via Harness Scaling Paper • 2608.15089 • Published 12 days ago • 441
SkillOpt: Executive Strategy for Self-Evolving Agent Skills Paper • 2605.23904 • Published May 22 • 264
VideoSeeker: Incentivizing Instance-level Video Understanding via Native Agentic Tool Invocation Paper • 2605.16079 • Published May 15 • 29
MMSkills: Towards Multimodal Skills for General Visual Agents Paper • 2605.13527 • Published May 14 • 123