AI & ML interests

Exact inference for robot policies. Proprio runs unmodified vision-language-action models (SmolVLA, pi0.5, GR00T) faster and proves every action chunk is bit-identical to the original model, with a verifiable certificate. Robotics inference serving, multi-robot batching, deterministic deployment on NVIDIA GPUs and Jetson.

Recent Activity

Gopal77021  updated a Space 7 days ago
ZenEdgeX/README
Gopal77021  updated a dataset 7 days ago
ZenEdgeX/proprio-certificates
Gopal77021  published a dataset 9 days ago
ZenEdgeX/proprio-certificates
View all activity

ZenEdgeX 's models

None public yet