StateM: Reaching 95.3% Raw Accuracy, or a \$15 Frontier Run, on Terminal-Bench 2.1 via Harness Scaling Paper • 2608.15089 • Published Aug 15 • 106
RESOURCE2SKILL: Distilling Executable Agent Skills from Human-Created Multimodal Resources Paper • 2606.29538 • Published Jul 16 • 90
SkillOpt: Executive Strategy for Self-Evolving Agent Skills Paper • 2605.23904 • Published May 22 • 221
Trust Your Critic: Robust Reward Modeling and Reinforcement Learning for Faithful Image Editing and Generation Paper • 2603.12247 • Published Mar 12 • 24
EvoTok: A Unified Image Tokenizer via Residual Latent Evolution for Visual Understanding and Generation Paper • 2603.12108 • Published Mar 12 • 8
Visionary: The World Model Carrier Built on WebGPU-Powered Gaussian Splatting Platform Paper • 2512.08478 • Published Dec 9, 2025 • 77
