The Teacher Is a Direction, Not a Destination: Extrapolating RL-Induced Representation Residuals in On-Policy Distillation Paper • 2609.36484 • Published 10 days ago • 582
MaLiang-Harness: A Programmable Path to Image and Video Generation Paper • 2609.34309 • Published 11 days ago • 413
Terminal-Universe: Turning Agent Trajectories into Scalable Terminal Environments Paper • 2609.04148 • Published Sep 3 • 249