GPT-Policy: Robots Learn New Tasks via VLM In-Context Learning
A new paper introduces GPT-Policy, a framework built on commercial VLMs that lets robots learn new tasks from in-context demonstrations without retraining or gradient updates. The approach enables robots to adapt to unfamiliar environments simply by prompting.
2026-09-17 ~ 2026-09-19 · 2 related posts
- GPT-Policy: In-Context Robot Learning with VLM Agents, No Gradient Updates — Dongzhou Cheng · 2026-09-17
- GPT-Policy teaches robots from human video demos via in-context learning, no retraining — KyeGomezB · 2026-09-19