mlsanityData to 14 Oct 2026

NeuroFaith: Evaluating LLM Self-Explanation Faithfulness via Internal Representation Alignment

arXiv 2506.09277 ↗First posted 10 Jun 2025cs.CLcited by 0

Not on the map. Why?

Abstract (arXiv)

Large Language Models (LLMs) can generate plausible free text self-explanations to justify their answers. However, these natural language explanations may not accurately reflect the model's actual reasoning process, pinpointing a lack of faithfulness. Existing faithfulness evaluation methods rely primarily on behavioral tests or computational block analysis without examining the semantic content of internal neural representations. This paper proposes NeuroFaith, a flexible framework that measures the faithfulness of LLM free text self-explanation by identifying key concepts within explanations and mechanistically testing whether these concepts actually influence the model's predictions. We show the versatility of NeuroFaith across 2-hop reasoning and classification tasks. Additionally, we develop a linear faithfulness probe based on NeuroFaith to detect unfaithful self-explanations from representation space and improve faithfulness through steering. NeuroFaith provides a principled approach to evaluating and enhancing the faithfulness of LLM free text self-explanations, addressing critical needs for trustworthy AI systems.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 52 of its 70 references that are in the map. First posted on arXiv 10 Jun 2025.

Builds on

52 of 70 references.

References · 52

  1. 2019PyTorch: An Imperative Style, High-Performance Deep Learning Librarycited by 20,305
  2. 2017A Unified Approach to Interpreting Model Predictionscited by 6,555
  3. 2016"Why Should I Trust You?": Explaining the Predictions of Any Classifiercited by 5,070
  4. 2017Axiomatic Attribution for Deep Networkscited by 3,384
  5. 2023Mistral 7Bcited by 3,164
  6. 2022Locating and Editing Factual Associations in GPTcited by 2,691
  7. 2020Shortcut Learning in Deep Neural Networkscited by 1,969
  8. 2024A Survey on LLM-as-a-Judgecited by 1,423
  9. 2023Representation Engineering: A Top-Down Approach to AI Transparencycited by 1,336
  10. 2017Interpretability Beyond Feature Attribution: Quantitative Testing with Concept Activation Vectors (TCAV)cited by 1,319
  11. 2022Interpretability in the Wild: a Circuit for Indirect Object Identification in GPT-2 smallcited by 1,165
  12. 2023Language Models Don't Always Say What They Think: Unfaithful Explanations in Chain-of-Thought Promptingcited by 1,122
38 more
  1. 2023Steering Llama 2 via Contrastive Activation Additioncited by 1,085
  2. 2022Toy Models of Superpositioncited by 971
  3. 2021MuSiQue: Multihop Questions via Single-hop Question Compositioncited by 916
  4. 2021Probing Classifiers: Promises, Shortcomings, and Advancescited by 876
  5. 2023Towards Automated Circuit Discovery for Mechanistic Interpretabilitycited by 724
  6. 2023The Linear Representation Hypothesis and the Geometry of Large Language Modelscited by 716
  7. 2024Phi-4 Technical Reportcited by 711
  8. 2023Measuring Faithfulness in Chain-of-Thought Reasoningcited by 703
  9. 2020Towards Faithfully Interpretable NLP Systems: How should we define and evaluate faithfulness?cited by 599
  10. 2024Mechanistic Interpretability for AI Safety -- A Reviewcited by 422
  11. 2023LEACE: Perfect linear concept erasure in closed formcited by 267
  12. 2024Rethinking Machine Unlearning for Large Language Modelscited by 257
  13. 2024Patchscopes: A Unifying Framework for Inspecting Hidden Representations of Language Modelscited by 235
  14. 2025AxBench: Steering LLMs? Even Simple Baselines Outperform Sparse Autoencoderscited by 218
  15. 2023Causal Abstraction: A Theoretical Foundation for Mechanistic Interpretabilitycited by 205
  16. 2024Transcoders Find Interpretable LLM Feature Circuitscited by 200
  17. 2020Measuring Association Between Labels and Free-Text Rationalescited by 160
  18. 2022Towards Faithful Model Explanation in NLP: A Surveycited by 130
  19. 2021Teach Me to Explain: A Review of Datasets for Explainable Natural Language Processingcited by 126
  20. 2024A Practical Review of Mechanistic Interpretability for Transformer-Based Language Modelscited by 118
  21. 2023Faithfulness Tests for Natural Language Explanationscited by 107
  22. 2024Are self-explanations from Large Language Models faithful?cited by 101
  23. 2024A Comprehensive Survey of Hallucination in Large Language, Image, Video and Audio Foundation Modelscited by 100
  24. 2024Information Flow Routes: Automatically Interpreting Language Models at Scalecited by 95
  25. 2023Can Large Language Models Explain Themselves? A Study of LLM-Generated Self-Explanationscited by 90
  26. 2024Hopping Too Late: Exploring the Limitations of Large Language Models on Multi-Hop Queriescited by 78
  27. 2023On Measuring Faithfulness or Self-consistency of Natural Language Explanationscited by 73
  28. 2024SelfIE: Self-Interpretation of Large Language Model Embeddingscited by 67
  29. 2022Multi-hop Question Answeringcited by 64
  30. 2025Measuring Chain of Thought Faithfulness by Unlearning Reasoning Stepscited by 47
  31. 2025Walk the Talk? Measuring the Faithfulness of Large Language Model Explanationscited by 45
  32. 2024The Probabilities Also Matter: A More Faithful Metric for Faithfulness of Free-Text Explanations in Large Language Modelscited by 29
  33. 2024Do Large Language Models Perform Latent Multi-Hop Reasoning without Exploiting Shortcuts?cited by 28
  34. 2024Position: Explain to Question not to Justifycited by 20
  35. 2024Towards Faithful Natural Language Explanations: A Study Using Activation Patching in Large Language Modelscited by 14
  36. 2025A Causal Lens for Evaluating Faithfulness Metricscited by 12
  37. 2024Self-AMPLIFY: Improving Small Language Models with Self Post Hoc Explanationscited by 8
  38. 2023Is Ignorance Bliss? The Role of Post Hoc Explanation Faithfulness and Alignment in Model Trust in Laypeople and Domain Expertscited by 7

50 most cited of 52 All 52 (JSON)

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026Training Large Language Models for Self-Explanation Faithfulness
  2. 2024Faithfulness vs. Plausibility: On the (Un)Reliability of Explanations from Large Language Models
  3. 2026A Positive Case for Faithfulness: LLM Self-Explanations Help Predict Model Behavior
  4. 2025Self-Critique and Refinement for Faithful Natural Language Explanations
  5. 2023Benchmarking Faithfulness: Towards Accurate Natural Language Explanations in Vision-Language Tasks
  6. 2026Can Large Language Models Still Explain Themselves? Investigating the Impact of Quantization on Self-Explanations
  7. 2024Enhancing Ethical Explanations of Large Language Models through Iterative Symbolic Refinement
  8. 2026Causally Grounded Mechanistic Interpretability for LLMs with Faithful Natural-Language Explanations