Uploaded August 2026 | Updated November 2026, 1 day ago
Max Simchowitz (Carnegie Mellon University)
https://simons.berkeley.edu/talks/max-simchowitz-carnegie-mellon-university-2026-08-07
Diffusion Generative Modeling: Progress and Next Steps
Recent years have seen tremendous progress in AI-driven robotics. These advancements have been attributed not only to increased data and improved hardware, but also to fundamental algorithmic innovations in the way robots are trained.
In this talk, we focus on one specific innovation: the introduction of ""generative control policies"" (GCPs), such as diffusion- and flow-based generative models as parameterizations of robot policies (policies being the mapping from observations to actions). The prevailing wisdom has been that GCPs enable robot policies to fit complex, often-multimodal action distributions that arise from diversely collected data. Through careful experimental and theoretical analysis, we suggest that GCPs routinely do not capture multi-modality in action data. Rather, we demonstrate that the increased performance afforded by GCPs are mediated through specific ""on-manifold"" inductive biases afforded by iterative computation and the incorporation of noise at training. In sum, our findings suggest that flow-based generative modeling offers algorithmic benefits unrelated to distribution matching, and point to the possibility of novel learning formalisms that achieve the same by more direct means.
Max Simchowitz (Carnegie Mellon University)
https://simons.berkeley.edu/talks/max-simchowitz-carnegie-mellon-university-2026-08-07
Diffusion Generative Modeling: Progress and Next Steps
Recent years have seen tremendous progress in AI-driven robotics. These advancements have been attributed not only to increased data and improved hardware, but also to fundamental algorithmic innovations in the way robots are trained.
In this talk, we focus on one specific innovation: the introduction of ""generative control policies"" (GCPs), such as diffusion- and flow-based generative models as parameterizations of robot policies (policies being the mapping from observations to actions). The prevailing wisdom has been that GCPs enable robot policies to fit complex, often-multimodal action distributions that arise from diversely collected data. Through careful experimental and theoretical analysis, we suggest that GCPs routinely do not capture multi-modality in action data. Rather, we demonstrate that the increased performance afforded by GCPs are mediated through specific ""on-manifold"" inductive biases afforded by iterative computation and the incorporation of noise at training. In sum, our findings suggest that flow-based generative modeling offers algorithmic benefits unrelated to distribution matching, and point to the possibility of novel learning formalisms that achieve the same by more direct means.










