Human-Alignment and Calibration of Inference-Time Uncertainty in Large Language Models

TL;DR

This paper evaluates how well inference-time uncertainty measures in large language models align with human uncertainty and traditional calibration metrics, highlighting measures that reflect human-like uncertainty and model correctness.

Contribution

It introduces novel evaluation methods for inference-time uncertainty and assesses their alignment with human uncertainty and model calibration.

Findings

Several uncertainty measures strongly align with human uncertainty.

Some measures demonstrate moderate to strong model calibration.

Alignment with human uncertainty does not necessarily imply preference alignment.

Abstract

There has been much recent interest in evaluating large language models for uncertainty calibration to facilitate model control and modulate user trust. Inference time uncertainty, which may provide a real-time signal to the model or external control modules, is particularly important for applying these concepts to improve LLM-user experience in practice. While many of the existing papers consider model calibration, comparatively little work has sought to evaluate how closely model uncertainty aligns to human uncertainty. In this work, we evaluate a collection of inference-time uncertainty measures, using both established metrics and novel variations, to determine how closely they align with both human group-level uncertainty and traditional notions of model calibration. We find that numerous measures show evidence of strong alignment to human uncertainty, even despite the lack of alignment to human answer preference. For those successful metrics, we find moderate to strong evidence of model calibration in terms of both correctness correlation and distributional analysis.