Energy Shields for Fairness
The paper titled 'Energy Shields for Fairness' introduces a novel approach to ensuring runtime fairness in decision-making processes. It presents energy shields as adaptive controllers that intervene probabilistically to maintain fairness over time. This method aims to provide both short-term safety and long-term liveness guarantees in fairness measures.
- ▪Energy shields are designed to monitor decision sequences and ensure fairness dynamically.
- ▪The intervention of energy shields is probabilistic, contrasting with traditional deterministic fairness shields.
- ▪The approach guarantees that fairness measures remain within specified target intervals over time.
Opening excerpt (first ~120 words) tap to expand
Computer Science > Artificial Intelligence arXiv:2605.24926 (cs) [Submitted on 24 May 2026] Title:Energy Shields for Fairness Authors:Filip Cano, Thomas A. Henzinger, Konstantin Kueffner View a PDF of the paper titled Energy Shields for Fairness, by Filip Cano and Thomas A. Henzinger and Konstantin Kueffner View PDF HTML (experimental) Abstract:Runtime fairness is not a one-time constraint but a dynamic property evaluated over a sequence of decisions. To ensure fairness at runtime, it is necessary to account for past decisions, information neglected by conventional, static classifiers. Traditional fairness shields enforce runtime fairness abruptly, by intervening \emph{deterministically} whenever a sequence of decisions violates the target for a running fairness measure.
…
Excerpt limited to ~120 words for fair-use compliance. The full article is at arXiv cs.AI.
Discussion
0 commentsMore from arXiv cs.AI
