July 24: Embodied Reasoning with World Models by Yilun Du

​Abstract: I'll present a couple methods showing how we can effectively reason with video models. I'll first talk about how we can directly reason with video models. I'll then talk about how video models can be combined with VLMs for hybrid reasoning and with policies for low level reasoning.

​Speaker: Yilun Du — Assistant Professor at Harvard University. Previously at MIT CSAIL, where his research focuses on generative models, world models, and embodied AI for robot learning and planning.

​Every week we pick one paper and go deep — video generation, world models, physical reasoning, diffusion, flow matching, and everything in between.

​Website: https://journal.video-reason.com/To join over zoom, please subscribe to get zoom link: Open Google Form

​Co-Hosted by 1943

​1943 is an invite-only community for researchers and builders who value intellectual depth. We host closed-door discussions and research talks in the Bay for those pushing the frontier. Follow us on X or Linkedin for future research talks and roundtable discussions!

​Co-Hosted by Moonlake AI

​Moonlake is a world model a simulation infrastructure platform to accelerate physical AI deployment. Follow them on X or Linkedin__.

​Co-Hosted by Tripo AI

​Tripo is an AI 3D generation platform that turns text and images into production-ready 3D models. Follow them on X or LinkedIn for the latest in generative 3D!