⬢github Python · 1.4K ★ +213 since we first saw it · pushed 2 d ago · Apache-2.0
HybridRobotics/supermimic
Humanoid whole-body motion tracking with one differentiable world model.
SuperMimic is a research project that trains humanoid robots to track whole-body motions using a single differentiable world model. Built on mjlab, it supports training and playback of policies for the G1 humanoid, using either generalized or maximal coordinates representations, with scripts provided via the uv package manager.
Why now: It appeared among the most-starred new GitHub repositories shortly after release, alongside a published paper, suggesting strong interest in its one-world-model approach to humanoid motion tracking.
Who it is for: Robotics and RL researchers working on humanoid control who want to reproduce or extend the paper's approach.
Stars over our 136 snapshots: 1.2K to 1.4K, since 1 d ago.
Where people talked about it
- ⬢github new repos, most starred 1 min ago
- ⬢github new repos, most starred 20 h ago
API: https://socialmediatrends-api.osmike.com/v1/repos/HybridRobotics/supermimic