TheMachine Press

A daily newspaper for the age of artificial intelligence.

Morning editionPermanent story

robotics

A Robot Resolved an Underspecified Task by Looking Around

CLUE tested hypotheses in the environment instead of assuming the goal and map were complete.

Published Updated Story ID: mp-2026-09-28-007
Read the complete editionStory JSON

Summary

CLUE tested hypotheses in the environment instead of assuming the goal and map were complete.

CLUE tackles natural-language tasks where a robot must infer both what success means and where relevant information might be. It forms task hypotheses, builds a language-embedded map online, tests possible plans through interaction and updates them as evidence arrives. On a Boston Dynamics Spot across 15 indoor and outdoor tasks, the authors report success within seven percentage points of an oracle and four times the rate of a planner without closed-loop feedback. The comparison covers three research environments and does not establish open-world reliability.

Why it matters

CLUE tested hypotheses in the environment instead of assuming the goal and map were complete.

Limits and context

  • The comparison covers three research environments and does not establish open-world reliability.

Key claims

  1. CLUE tested hypotheses in the environment instead of assuming the goal and map were complete.

    Qualification: The comparison covers three research environments and does not establish open-world reliability.

    Evidence: source-2026-09-28-007

Sources

  1. arXiv preprint 2609.30428arXiv · primary research

Corrections

No corrections have been recorded for this story.