Mapping Clinical Doubt: Locating Linguistic Uncertainty in LLMs

TL;DR

This paper investigates how large language models internally represent linguistic uncertainty in clinical texts, revealing depth-dependent sensitivity that enhances understanding of their interpretability and reliability in medical applications.

Contribution

The study introduces a novel probing metric, MSU, to analyze layerwise sensitivity to uncertainty cues in LLMs, and demonstrates structured depth-dependent encoding of epistemic information.

Findings

LLMs show structured sensitivity to clinical uncertainty.

Sensitivity to uncertainty increases in deeper layers.

Epistemic information is encoded progressively in LLMs.

Abstract

Large Language Models (LLMs) are increasingly used in clinical settings, where sensitivity to linguistic uncertainty can influence diagnostic interpretation and decision-making. Yet little is known about where such epistemic cues are internally represented within these models. Distinct from uncertainty quantification, which measures output confidence, this work examines input-side representational sensitivity to linguistic uncertainty in medical text. We curate a contrastive dataset of clinical statements varying in epistemic modality (e.g., 'is consistent with' vs. 'may be consistent with') and propose Model Sensitivity to Uncertainty (MSU), a layerwise probing metric that quantifies activation-level shifts induced by uncertainty cues. Our results show that LLMs exhibit structured, depth-dependent sensitivity to clinical uncertainty, suggesting that epistemic information is progressively encoded in deeper layers. These findings reveal how linguistic uncertainty is internally represented in LLMs, offering insight into their interpretability and epistemic reliability.