1/9 Current optimizers are static during training. What if the optimizer could think and react while the network trains? We let an agent watch the training dynamics and rewrite the optimizer's source code as training goes: a reactive optimizer. In every language and image setting we tried that has no tuned recipe, it matches or beats the best recipe one can one-shot, and it guards against loss spikes and overfitting by itself. In the near future, optimization decisions may belong to agents, not humans. 📝 xingjianbai.com/reactive-opt… 💻 Try it on your task: github.com/xingjian-bai/opal 🧵

This work wouldn't have been possible without support from the @AMD University Program thanks to @secunit64. Our video planner was trained on as many as 32 AMD MI350X GPUs!! Our feedback has led to FlexAttention speed ups that have already been merged into the @PyTorch nightly!

Do coding agents understand the dynamics of the world well enough to reconstruct it from video? We introduce 4DCodeBench to evaluate this ability through 4D inverse graphics. 4DCodeBench.com/ <🧵>

I am open-sourcing DeckWerk, a multi-platform slide editor with stellar video support, real-time collaboration, and Agent support. I used to be on Linux but had to switch to Mac b/c there was no good enough slide editor - @OmarchyLinux & @dhh inspired me to vibe-code my own, so I'm finally back on Linux! Install from source via GitHub (installers coming soon!): github.com/vsitzmann/deckwer…

We discovered that our latest Dataset Distillation project can be used to create some beautiful synthetic images based on an artist's body of work! Come see us at the @eccvconf Art Gallery starting today! Explanation and some of my favorites in thread below: 1/ (Claude Monet)

Introducing MilliVid, our new method for long-context video generation! MilliVid creates videos that are consistent over long time spans, without using retrieval heuristics or 3D maps! (1/n) davidcharatan.com/millivid/#

Many of my students / collaborators are at CVPR - find them & chat! @ottogin1, diffusion models @ericmchen1, latent actions & robotics @RyuHyunwoooo, latent actions & robotics @ekim2339, robotics, view synthesis @SimulatedAnneal, robotics @twmitchel, latent actions & video

How well can a model watch a short video of some physical dynamics and actually predict what happens next? Introducing MPMWorlds: a new dataset and benchmark to evaluate how well models can reconstruct and extrapolate physical dynamics from video. zzigak.github.io/mpmworlds/ 🧵👇 (1/n)

Come check out our talk at #CHI2026 tomorrow (Fri) at 9:45am! We present CineCraft, an interactive mobile app that unifies planning, capture, and post-processing for cinematography on a single device. Fun combining my filmmaking hobbies with research! megatran.github.io/cinecraft