Navigating the Grey Area: How Expressions of Uncertainty and Overconfidence Affect Language Models
Submission Type: Regular Long Paper
Submission Track: Language Modeling and Analysis of Language Models
Submission Track 2: Theme Track: Large Language Models and the Future of NLP
Keywords: expressions of uncertainty, analysis of language models
TL;DR: Expressions of certainty in prompts hurt the accuracy of language models and we trace this effect back to pretraining datasets.
Abstract: The increased deployment of LMs for real-world tasks involving knowledge and facts makes it important to understand model epistemology: what LMs think they know, and how their attitudes toward that knowledge are affected by language use in their inputs. Here, we study an aspect of model epistemology: how epistemic markers of certainty, uncertainty, or evidentiality like "I'm sure it's", "I think it's", or "Wikipedia says it's" affect models, and whether they contribute to model failures. We develop a typology of epistemic markers and inject 50 markers into prompts for question answering. We find that LMs are highly sensitive to epistemic markers in prompts, with accuracies varying more than 80%. Surprisingly, we find that expressions of high certainty result in a 7% decrease in accuracy as compared to low certainty expressions; similarly, factive verbs hurt performance, while evidentials benefit performance. Our analysis of a popular pretraining dataset shows that these markers of uncertainty are associated with answers on question-answering websites, while markers of certainty are associated with questions. These associations may suggest that the behavior of LMs is based on mimicking observed language use, rather than truly reflecting epistemic uncertainty.
Submission Number: 809
Loading