AI RESEARCH
Proper Scoring Rules for Agentic Uncertainty Quantification
arXiv CS.AI
•
ArXi:2605.24756v1 Announce Type: new Language-model agents increasingly emit uncertainty signals throughout a trajectory, but existing agentic UQ evaluations often conflate ranking usefulness with probabilistic truthfulness. AUROC, AUPRC, risk-coverage, Trajectory ECE, and scalarized trajectory scores evaluate discrimination, binwise calibration, or collapsed summaries, but do not strictly elicit the full prefix-conditioned success-probability trace $q_t = P^{\pi}(Y=1 | H_t)$. Building on prequential proper scoring, we.