Factual Confidence of LLMs: on Reliability and Robustness of Current Estimators

Matéo Mahaut, Laura Aina, Paula Czarnowska, Momchil Hardalov, Thomas Müller, Lluis Marquez


Abstract
Large Language Models (LLMs) tend to be unreliable on fact-based answers.To address this problem, NLP researchers have proposed a range of techniques to estimate LLM’s confidence over facts. However, due to the lack of a systematic comparison, it is not clear how the different methods compare to one other.To fill this gap, we present a rigorous survey and empirical comparison of estimators of factual confidence.We define an experimental framework allowing for fair comparison, covering both fact-verification and QA. Our experiments across a series of LLMs indicate that trained hidden-state probes provide the most reliable confidence estimates; albeit at the expense of requiring access to weights and supervision data. We also conduct a deeper assessment of the methods, in which we measure the consistency of model behavior under meaning-preserving variations in the input. We find that the factual confidence of LLMs is often unstable across semantically equivalent inputs, suggesting there is much room for improvement for the stability of models’ parametric knowledge.
Anthology ID:
2024.acl-long.250
Volume:
Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)
Month:
August
Year:
2024
Address:
Bangkok, Thailand
Editors:
Lun-Wei Ku, Andre Martins, Vivek Srikumar
Venue:
ACL
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
4554–4570
Language:
URL:
https://aclanthology.org/2024.acl-long.250
DOI:
10.18653/v1/2024.acl-long.250
Bibkey:
Cite (ACL):
Matéo Mahaut, Laura Aina, Paula Czarnowska, Momchil Hardalov, Thomas Müller, and Lluis Marquez. 2024. Factual Confidence of LLMs: on Reliability and Robustness of Current Estimators. In Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 4554–4570, Bangkok, Thailand. Association for Computational Linguistics.
Cite (Informal):
Factual Confidence of LLMs: on Reliability and Robustness of Current Estimators (Mahaut et al., ACL 2024)
Copy Citation:
PDF:
https://aclanthology.org/2024.acl-long.250.pdf