NeoHorse-1: Towards Recursive Self-Improvement via Agentic Post-Training with Routing Harness Paper • 2609.08183 • Published 27 days ago • 327
RoboTok: An Internet-Scale Data Engine for Human Demonstration Retrieval and Dexterous Manipulation Learning Paper • 2609.03199 • Published Sep 2 • 31
Agentic Game Development as a Verifiable Trajectory Data Engine for Scaling World Models Paper • 2608.25518 • Published Aug 26 • 60
DistillAlign: Coordinating Mode Covering and Mode Seeking in Autoregressive Video Distillation Paper • 2607.26811 • Published Jul 29 • 18