Wendi Chen

Real-world robot policies can learn autonomously from sparse rewards by treating value predictions as task-progress memory, outperforming offline training by $32.8\%$ success rate.

Robots can now evolve their own code, learning from task execution to autonomously repair and reuse skills, achieving superior performance on complex manipulation benchmarks.

FA-RDP achieves superior success rates in contact-rich manipulation while preserving diverse action modes, revolutionizing how we approach multimodal decision-making in robotics.