Agentic Game Development as a Verifiable Trajectory Data Engine for Scaling World Models Paper • 2608.25518 • Published 7 days ago • 190
HarnessEval-W: Agentifying the Evaluation of Visual Worlds Paper • 2608.16859 • Published 16 days ago • 339
logan7000/cogrpo-n3-strict-qwen25-3b-x-llama32-3b-x-qwen3-1p7b-math345-groupC-qwen3-best Reinforcement Learning • 2B • Updated 21 days ago • 17 • 1
TARS: Timestep-Aware Data Scaling for 3D-Free Video Re-Shooting Paper • 2607.28261 • Published Jul 30 • 116
Qwen-UI-Agent Technical Report: Toward Next-Generation Real-World Centric Foundation GUI Agents Paper • 2607.28227 • Published Jul 30 • 310
Read It Back: Pretrained MLLMs Are Zero-Shot Reward Models for Text-to-Image Generation Paper • 2607.11886 • Published Jul 13 • 86
PixWorld: Unifying 3D Scene Generation and Reconstruction in Pixel Space Paper • 2607.05373 • Published Jul 6 • 66
BioInsight: Multi-Agent Orchestration for Interactive Biomedical Knowledge Discovery Paper • 2606.20997 • Published Jun 19 • 13