On Measuring Faithfulness or Self-consistency of Natural Language Explanations

TL;DR

This paper argues that current faithfulness tests for LLM explanations actually measure self-consistency, introduces a new benchmark and a fine-grained measure CC-SHAP to better assess model faithfulness.

Contribution

It clarifies the distinction between faithfulness and self-consistency, and introduces CC-SHAP, a detailed measure for analyzing LLM self-consistency and explanation faithfulness.

Findings

Existing tests mainly measure self-consistency, not true faithfulness.

The Comparative Consistency Bank compares tests across 11 LLMs and 5 tasks.

CC-SHAP provides a detailed analysis of input contribution to answers and explanations.

Abstract

Large language models (LLMs) can explain their predictions through post-hoc or Chain-of-Thought (CoT) explanations. But an LLM could make up reasonably sounding explanations that are unfaithful to its underlying reasoning. Recent work has designed tests that aim to judge the faithfulness of post-hoc or CoT explanations. In this work we argue that these faithfulness tests do not measure faithfulness to the models' inner workings -- but rather their self-consistency at output level. Our contributions are three-fold: i) We clarify the status of faithfulness tests in view of model explainability, characterising them as self-consistency tests instead. This assessment we underline by ii) constructing a Comparative Consistency Bank for self-consistency tests that for the first time compares existing tests on a common suite of 11 open LLMs and 5 tasks -- including iii) our new self-consistency measure CC-SHAP. CC-SHAP is a fine-grained measure (not a test) of LLM self-consistency. It compares how a model's input contributes to the predicted answer and to generating the explanation. Our fine-grained CC-SHAP metric allows us iii) to compare LLM behaviour when making predictions and to analyse the effect of other consistency tests at a deeper level, which takes us one step further towards measuring faithfulness by bringing us closer to the internals of the model than strictly surface output-oriented tests. Our code is available at \url{https://github.com/Heidelberg-NLP/CC-SHAP}