Takuma Seno

SimBa: Simplicity Bias for Scaling Up Parameters in Deep Reinforcement Learning

Hyperspherical Normalization for Scalable Deep Reinforcement Learning