#artificial intelligence
Apr 2026
Interval POMDP Shielding for Imperfect-Perception Agents
This work designs a runtime shield that comes with a finite-horizon guarantee: with high probability over the training data, if the true perception uncertainty rates lie within the learned intervals, then every action admitted by the shield satisfies a stated lower bound on safety.
William Scarbro, Ravi Mangal
· arXiv.org · 0 citations