Conformal Reachability for Safe Control in Unknown Environments Conference: Proc. of the 25th International Conference on Autonomous Agents and Multiagent Systems (AAMAS 2026); May 25 – 29, 2026; Paphos, Cyprus 115
Xinhang Ma, Junlin Wu, Yiannis Kantaros, Yevgeniy Vorobeychik · 2026
Read Conformal Reachability for Safe Control in Unknown Environments Conference: Proc. of the 25th International Conference on Autonomous Agents and Multiagent Systems (AAMAS 2026); May 25 – 29, 2026; Paphos, Cyprus 115 by Xinhang Ma, Junlin Wu, Yiannis Kantaros, Yevgeniy Vorobeychik (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-2602-03799.txt.
Description
A controller for a system nobody has modeled exactly cannot lean on the usual safety proofs. Xinhang Ma and 3 co-authors combine conformal prediction with reachability analysis in a 2026 paper to certify safety anyway. Conformal prediction supplies valid uncertainty intervals for the unknown dynamics at each step, from data alone.
Reachability analysis then checks whether safety holds throughout those uncertainty bounds. The authors also give an algorithm that trains control policies to earn high reward while stretching the horizon over which the probabilistic guarantee holds. Seven test settings across four domains, from cartpole to drones, cover affine and nonlinear safety rules. Their policies achieve the strongest provable safety guarantees while keeping average reward high.
Catalog record
| CREATOR | Xinhang Ma, Junlin Wu, Yiannis Kantaros, Yevgeniy Vorobeychik |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 55 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2602-03799