Human-Alignment, Calibration, and Activation Patterns in Large Language Model Uncertainty

ArXi:2605.30675v1 Announce Type: cross Uncertainty Quantification is a large and growing subfield of large language model behavioral analysis. Primarily to recognize and combat hallucination, the field has largely focused on measuring and improving calibration, the accuracy of uncertainty judgments to task efficacy. In this work, we investigate the relatively underexplored question of how similar large language model uncertainty is to human uncertainty.