# Towards Faithful Natural Language Explanations: A Study Using Activation Patching in Large Language Models

arXiv: https://arxiv.org/abs/2410.14155 · HTML: https://mlsanity.com/paper/2410.14155/

## Key facts

- Cited by 14 papers in the map (data through 14 Oct 2026).
- Placed in Knowledge and representations inside LLMs, one of 13 directions in Mechanistic interpretability.
- Builds on 28 of its 45 references that are in the map.
- First posted on arXiv 18 Oct 2024.

## Placed in

- Mechanistic interpretability / [Knowledge and representations inside LLMs](https://mlsanity.com/mechinterp/knowledge-and-representations-inside-llms/)

## Builds on (most cited first, 28 of 28)

- [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017)
- [GPT-4 Technical Report](https://mlsanity.com/paper/2303.08774/) (2023)
- [A Unified Approach to Interpreting Model Predictions](https://mlsanity.com/paper/1705.07874/) (2017)
- [Locating and Editing Factual Associations in GPT](https://mlsanity.com/paper/2202.05262/) (2022)
- [Language Models Don't Always Say What They Think: Unfaithful Explanations in Chain-of-Thought Prompting](https://mlsanity.com/paper/2305.04388/) (2023)
- [Progress measures for grokking via mechanistic interpretability](https://mlsanity.com/paper/2301.05217/) (2023)
- [Measuring Faithfulness in Chain-of-Thought Reasoning](https://mlsanity.com/paper/2307.13702/) (2023)
- [Towards Faithfully Interpretable NLP Systems: How should we define and evaluate faithfulness?](https://mlsanity.com/paper/2004.03685/) (2020)
- [ERASER: A Benchmark to Evaluate Rationalized NLP Models](https://mlsanity.com/paper/1911.03429/) (2019)
- [e-SNLI: Natural Language Inference with Natural Language Explanations](https://mlsanity.com/paper/1812.01193/) (2018)
- [Explain Yourself! Leveraging Language Models for Commonsense Reasoning](https://mlsanity.com/paper/1906.02361/) (2019)
- [Towards Best Practices of Activation Patching in Language Models: Metrics and Methods](https://mlsanity.com/paper/2309.16042/) (2023)
- [Rationalizing Neural Predictions](https://mlsanity.com/paper/1606.04155/) (2016)
- [Causal Mediation Analysis for Interpreting Neural NLP: The Case of Gender Bias](https://mlsanity.com/paper/2004.12265/) (2020)
- [Measuring Association Between Labels and Free-Text Rationales](https://mlsanity.com/paper/2010.12762/) (2020)
- [Making Reasoning Matter: Measuring and Improving Faithfulness of Chain-of-Thought Reasoning](https://mlsanity.com/paper/2402.13950/) (2024)
- [Faithfulness Tests for Natural Language Explanations](https://mlsanity.com/paper/2305.18029/) (2023)
- [Faithfulness vs. Plausibility: On the (Un)Reliability of Explanations from Large Language Models](https://mlsanity.com/paper/2402.04614/) (2024)
- [The Out-of-Distribution Problem in Explainability and Search Methods for Feature Importance Explanations](https://mlsanity.com/paper/2106.00786/) (2021)
- [Aligning Faithful Interpretations with their Social Attribution](https://mlsanity.com/paper/2006.01067/) (2020)
- [GPTEval: A Survey on Assessments of ChatGPT and GPT-4](https://mlsanity.com/paper/2308.12488/) (2023)
- [Faithful Explanations of Black-box NLP Models Using LLM-generated Counterfactuals](https://mlsanity.com/paper/2310.00603/) (2023)
- [SemEval-2020 Task 4: Commonsense Validation and Explanation](https://mlsanity.com/paper/2007.00236/) (2020)
- [The Probabilities Also Matter: A More Faithful Metric for Faithfulness of Free-Text Explanations in Large Language Models](https://mlsanity.com/paper/2404.03189/) (2024)
- [A Comprehensive Review on Financial Explainable AI](https://mlsanity.com/paper/2309.11960/) (2023)
- [Explainable Natural Language Processing for Corporate Sustainability Analysis](https://mlsanity.com/paper/2407.17487/) (2024)
- [Self-training Large Language Models through Knowledge Detection](https://mlsanity.com/paper/2406.11275/) (2024)
- [Plausible Extractive Rationalization through Semi-Supervised Entailment Signal](https://mlsanity.com/paper/2402.08479/) (2024)

## Built on by (most cited first, 14 of 14)

- [Locate, Steer, and Improve: A Practical Survey of Actionable Mechanistic Interpretability in Large Language Models](https://mlsanity.com/paper/2601.14004/) (2026)
- [How does Chain of Thought Think? Mechanistic Interpretability of Chain-of-Thought Reasoning with Sparse Autoencoding](https://mlsanity.com/paper/2507.22928/) (2025)
- [A Positive Case for Faithfulness: LLM Self-Explanations Help Predict Model Behavior](https://mlsanity.com/paper/2602.02639/) (2026)
- [Debiasing CLIP: Interpreting and Correcting Bias in Attention Heads](https://mlsanity.com/paper/2505.17425/) (2025)
- [SafeSeek: Universal Attribution of Safety Circuits in Language Models](https://mlsanity.com/paper/2603.23268/) (2026)
- [Counting Circuits: Mechanistic Interpretability of Visual Reasoning in Large Vision-Language Models](https://mlsanity.com/paper/2603.18523/) (2026)
- [Faithfulness Metrics Don't Measure Faithfulness: A Meta-Evaluation with Ground Truth](https://mlsanity.com/paper/2605.25052/) (2026)
- [How LLMs Are Persuaded: A Few Attention Heads, Rerouted](https://mlsanity.com/paper/2605.09314/) (2026)
- [How Chain-of-Thought Works? Tracing Information Flow from Decoding, Projection, and Activation](https://mlsanity.com/paper/2507.20758/) (2025)
- [Mechanistic Interpretability of Chain-of-Thought Reasoning via Sequential Activation Patching](https://mlsanity.com/paper/2608.22332/) (2026)
- [Silence Is Endorsement: Verification-Status Laundering in LLM Agent Pipelines](https://mlsanity.com/paper/2609.20211/) (2026)
- [NeuroFaith: Evaluating LLM Self-Explanation Faithfulness via Internal Representation Alignment](https://mlsanity.com/paper/2506.09277/) (2025)
- [Causally Grounded Mechanistic Interpretability for LLMs with Faithful Natural-Language Explanations](https://mlsanity.com/paper/2603.09988/) (2026)
- [What Would Falsify It? A Variable Specific Evidence Standard for Mechanistic Claims About Self Explanation](https://mlsanity.com/paper/2609.32670/) (2026)

## Abstract (arXiv)

Large Language Models (LLMs) are capable of generating persuasive Natural Language Explanations (NLEs) to justify their answers. However, the faithfulness of these explanations should not be readily trusted at face value. Recent studies have proposed various methods to measure the faithfulness of NLEs, typically by inserting perturbations at the explanation or feature level. We argue that these approaches are neither comprehensive nor correctly designed according to the established definition of faithfulness. Moreover, we highlight the risks of grounding faithfulness findings on out-of-distribution samples. In this work, we leverage a causal mediation technique called activation patching, to measure the faithfulness of an explanation towards supporting the explained answer. Our proposed metric, Causal Faithfulness quantifies the consistency of causal attributions between explanations and the corresponding model outputs as the indicator of faithfulness. We experimented across models varying from 2B to 27B parameters and found that models that underwent alignment tuning tend to produce more faithful and plausible explanations. We find that Causal Faithfulness is a promising improvement over existing faithfulness tests by taking into account the model's internal computations and avoiding out of distribution concerns that could otherwise undermine the validity of faithfulness assessments. We release the code in urlhttps://github.com/wj210/Causal-Faithfulness