RoboCritics: Enabling Reliable End-to-End LLM Robot Programming through Expert-Informed Critics DOI: 10.1145/3757279.3785550 ISBN: 979-8-4007-2128-1/2026/03 Conference: Proceedings of the 21st ACM/IEEE International Conference on Human-Robot Interaction; March 16–19, 2026; Edinburgh, Scotland, UK Proceedings of the 21st ACM/IEEE International Conference on Human-Robot Interaction (HRI ’26), March 16–19, 2026, Edinburgh, Scotland, UK CCS: Human-centered computing Interaction paradigms CCS: Computer systems organization External interfaces for robotics
Callie Y. Kim, Nathan Thomas White, Evan He, Frederic Sala, Bilge Mutlu · 2026
Read RoboCritics: Enabling Reliable End-to-End LLM Robot Programming through Expert-Informed Critics DOI: 10.1145/3757279.3785550 ISBN: 979-8-4007-2128-1/2026/03 Conference: Proceedings of the 21st ACM/IEEE International Conference on Human-Robot Interaction; March 16–19, 2026; Edinburgh, Scotland, UK Proceedings of the 21st ACM/IEEE International Conference on Human-Robot Interaction (HRI ’26), March 16–19, 2026, Edinburgh, Scotland, UK CCS: Human-centered computing Interaction paradigms CCS: Computer systems organization External interfaces for robotics by Callie Y. Kim, Nathan Thomas White, Evan He, Frederic Sala, Bilge Mutlu (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2603-06842.txt.
Description
Kim and colleagues add expert motion critics to LLM-based robot coding for users. Novices struggle to re-task robots without specialist skill, and generated code is opaque code, hard to verify or debug, with real safety and trust risks. Critics encode robotics know-how and check motion traces for fast joints, collisions, and unsafe end-effector poses. When a problem appears, critics show clear feedback and offer one-click fixes that send clear messages back to the model. The loop refines code steadily while keeping users in charge.
Catalog record
| CREATOR | Callie Y. Kim, Nathan Thomas White, Evan He, Frederic Sala, Bilge Mutlu |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 45198 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2603.06842