Kevin Black(@kvablack):This is great work! And I finally get to live my dream of Schmidhubering, just once: Using the average of rollouts from the same state as a baseline is exactly what I did in the original DDPO paper (which predates GRPO). But I thought it was boring so I put it in the Appendix :)

Kevin Black

@kvablack

phd @berkeley_ai, research @physical_int

Joined March 2018

133 Following 3.4K Followers

Kevin Black@kvablack

2026.06.30 19:36

This is great work! And I finally get to live my dream of Schmidhubering, just once: Using the average of rollouts from the same state as a baseline is exactly what I did in the original DDPO paper (which predates GRPO). But I thought it was boring so I put it in the Appendix :)