Rescue trajectories over real alpine terrain (Chamonix, Zermatt, Aosta, Dolomites) and tiny MLP policies.
Edgar Moreau
Ethgar
P(doom) Rejects the premise
AI & ML interests
Autonomous UAVs, Embodied AI, World models, application over research.
Recent Activity
updated a model about 20 hours ago
Ethgar/stringman-world-model published a model about 20 hours ago
Ethgar/stringman-world-model updated a Space 2 days ago
Ethgar/robot-lab-notebookOrganizations
Rescue trajectories over real alpine terrain (Chamonix, Zermatt, Aosta, Dolomites) and tiny MLP policies.
G1 policies trained in mjlab on HF Jobs (~$1 each), measured for energy: walk/stand, carry from PRISM clips.
Clean-room SkyJEPA reimplementation: a 5.6k-parameter JEPA world model flown by MPPI, its data, and the v2 that was not promoted.
Real SO-101 arm: teleop datasets, a measured desk twin, ACT/SmolVLA evaluated in sim, then live MPPI grasps on the real arm.