Stars
Official loco-manipulation simulation benchmark environments from the HumanoidMimicGen project.
Benchmarking Coding Agents for Interactive World Modeling from Embodied Videos - Rethinking Embodied Real2Sim from a Software Engineering Perspective
Project page: Workhorse, learning robust whole-body humanoid loco-manipulation from human data
Official implementation of FlowHMR, a generative model for physically plausible motion capture from monocular video.
4DCodeBench: Benchmarking Agents on Inverse Graphics of Dynamic Scenes
SONIC adpation for new full size humanoid agibot A3！
[NeurIPS 2026] Beyond Data Scaling: Representation-Centric Continued Pre-training for Vision-Language-Action Models
[IROS 2026] ULTRA: Unified Multimodal Control for Autonomous Humanoid Whole-Body Loco-Manipulation
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.
RoboCasa: Large-Scale Simulation of Everyday Tasks for Generalist Robots
EmbodiedJev: MuJoCo robot decision workbench with MiniCPM5-2B, Jev and compatible model APIs
RPent: Agentic Infrastructure for the Physical World
Official repository for "OpenWAM: An Open, Modular Exploration Towards Systematic World–Action Model Pretraining".
ENPIRE is an agentic framework that allows AI coding agents to autonomously self-improve real-world robot policies through closed-loop physical trial, error, and evolution
GPT-6 Astra for embodied AI and robotics.
VideoDeltaNet-H3: Live T2VA / I2VA / FL2VA / Ref2VA-like generation based on Minimax H3.
A skill to stop your coding agent from burying the answer. ADHD-friendly output.
Open source evals for physical AI. Run any LLM/VLA on any arm/humanoid against any real/sim benchmark.
DriveZero: End-to-End Driving Beyond Human Demonstrations
The official codebase of Unified Motion Retargeting
Building General-Purpose Robots Based on Embodied Foundation Model