Evaluating Readability and Faithfulness of Concept-based Explanations

Meng Li, Haoran Jin, Ruixuan Huang, Zhihao Xu, Defu Lian, Zijia Lin, Di Zhang, Xiting Wang


Abstract
With the growing popularity of general-purpose Large Language Models (LLMs), comes a need for more global explanations of model behaviors. Concept-based explanations arise as a promising avenue for explaining high-level patterns learned by LLMs. Yet their evaluation poses unique challenges, especially due to their non-local nature and high dimensional representation in a model’s hidden space. Current methods approach concepts from different perspectives, lacking a unified formalization. This makes evaluating the core measures of concepts, namely faithfulness or readability, challenging. To bridge the gap, we introduce a formal definition of concepts generalizing to diverse concept-based explanations’ settings. Based on this, we quantify the faithfulness of a concept explanation via perturbation. We ensure adequate perturbation in the high-dimensional space for different concepts via an optimization problem. Readability is approximated via an automatic and deterministic measure, quantifying the coherence of patterns that maximally activate a concept while aligning with human understanding. Finally, based on measurement theory, we apply a meta-evaluation method for evaluating these measures, generalizable to other types of explanations or tasks as well. Extensive experimental analysis has been conducted to inform the selection of explanation evaluation measures.
Anthology ID:
2024.emnlp-main.36
Volume:
Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing
Month:
November
Year:
2024
Address:
Miami, Florida, USA
Editors:
Yaser Al-Onaizan, Mohit Bansal, Yun-Nung Chen
Venue:
EMNLP
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
607–625
Language:
URL:
https://aclanthology.org/2024.emnlp-main.36
DOI:
Bibkey:
Cite (ACL):
Meng Li, Haoran Jin, Ruixuan Huang, Zhihao Xu, Defu Lian, Zijia Lin, Di Zhang, and Xiting Wang. 2024. Evaluating Readability and Faithfulness of Concept-based Explanations. In Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing, pages 607–625, Miami, Florida, USA. Association for Computational Linguistics.
Cite (Informal):
Evaluating Readability and Faithfulness of Concept-based Explanations (Li et al., EMNLP 2024)
Copy Citation:
PDF:
https://aclanthology.org/2024.emnlp-main.36.pdf