CMU-IntentLab/StressDream

Official implementation of Video Prediction Policy 2: Predict Better, Act Better

Code for "Evaluating Robot Policies in a World Model".

Code to pretrain, fine-tune, and evaluate DreamZero and run sim & real-world evals

Dream.exe: Can Video Generation Models Dream Executable Robot Manipulation?

Empower steerable world action modeling with reasoning-augmented memory and event-grounded pretraining under multimodal control for in-context learning.

[RSS 2023] Diffusion Policy Visuomotor Policy Learning via Action Diffusion

OSS of roboJEPA

FLUX 3 Action: open weights 7B world action model from Black Forest Labs. Train, fine-tune and run action prediction for robots (DROID, SO-101), simulators and games.

Robometer: Scaling General-Purpose Robotic Reward Models via Trajectory Comparisons