Zilin Kang

Entropy Regularizing Activation: Boosting Continuous Control, Large Language Models, and Image Classification with Activation as Entropy Constraints

A Forget-and-Grow Strategy for Deep Reinforcement Learning Scaling in Continuous Control