MMODELYST
People/Zeyu Huang
PPL

Zeyu Huang

Papers (2)
Context Training with Active Information SeekingThe Cancellation Hypothesis in Critic-Free RL: From Outcome Rewards to Token Credits