FastTD3: Simple, Fast, and Capable Reinforcement Learning for Humanoid Control Paper • 2505.22642 • Published May 28, 2025 • 4