LLMs help robots understand vague instructions and focus on key details

To help robots do chores in places like homes and factories, a new approach from MIT uses one language model to clarify users’ instructions, then another to ignore irrelevant info.

Published: 2026-06-26