mlsanityData to 7 Oct 2026

Faithfulness Tests for Natural Language Explanations

arXiv 2305.18029 ↗First posted 29 May 2023cs.CL cs.AIcited by 107

Not on the map. Why?

Cited by 29 in LLM reasoning, but placed by its own abstract, which didn’t fit.

Abstract (arXiv)

Explanations of neural models aim to reveal a model's decision-making process for its predictions. However, recent work shows that current methods giving explanations such as saliency maps or counterfactuals can be misleading, as they are prone to present reasons that are unfaithful to the model's inner workings. This work explores the challenging question of evaluating the faithfulness of natural language explanations (NLEs). To this end, we present two tests. First, we propose a counterfactual input editor for inserting reasons that lead to counterfactual predictions but are not reflected by the NLEs. Second, we reconstruct inputs from the reasons stated in the generated NLEs and check how often they lead to the same predictions. Our tests can evaluate emerging NLE models, proving a fundamental tool in the development of faithful NLEs.

Key facts. Cited by 107 papers in the map (data through 7 Oct 2026). Builds on 36 of its 47 references that are in the map. First posted on arXiv 29 May 2023.

Builds on

36 of 47 references.

References · 36

  1. 2014Adam: A Method for Stochastic Optimizationcited by 36,960
  2. 2019Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformercited by 14,491
  3. 2015A large annotated corpus for learning natural language inferencecited by 1,837
  4. 2017Explanation in Artificial Intelligence: Insights from the Social Sciencescited by 1,201
  5. 2018Sanity Checks for Saliency Mapscited by 1,149
  6. 2019ERASER: A Benchmark to Evaluate Rationalized NLP Modelscited by 523
  7. 2018e-SNLI: Natural Language Inference with Natural Language Explanationscited by 468
  8. 2019Fooling LIME and SHAP: Adversarial Attacks on Post hoc Explanation Methodscited by 440
  9. 2019Explain Yourself! Leveraging Language Models for Commonsense Reasoningcited by 367
  10. 2021Polyjuice: Generating Counterfactuals for Explaining, Evaluating, and Improving Modelscited by 219
  11. 2019Explanations can be manipulated and geometry is to blamecited by 215
  12. 2020Measuring Association Between Labels and Free-Text Rationalescited by 159
24 more
  1. 2020A Diagnostic Study of Explainability Techniques for Text Classificationcited by 158
  2. 2020WT5?! Training Text-to-Text Models to Explain their Predictionscited by 148
  3. 2020Generating Fact Checking Explanationscited by 132
  4. 2021Teach Me to Explain: A Review of Datasets for Explainable Natural Language Processingcited by 126
  5. 2021Counterfactual Explanations Can Be Manipulatedcited by 113
  6. 2020Explaining NLP Models via Minimal Contrastive Editing (MiCE)cited by 105
  7. 2019Rethinking Cooperative Rationalization: Introspective Extraction and Complement Controlcited by 96
  8. 2020Leakage-Adjusted Simulatability: Can Models Generate Non-Trivial Explanations of Their Behavior in Natural Language?cited by 95
  9. 2021e-ViL: A Dataset and Benchmark for Natural Language Explanations in Vision-Language Taskscited by 84
  10. 2021Few-Shot Self-Rationalization with Natural Language Promptscited by 83
  11. 2021Contrastive Explanations for Model Interpretabilitycited by 75
  12. 2019Make Up Your Mind! Adversarial Generation of Inconsistent Natural Language Explanationscited by 71
  13. 2020Fairwashing Explanations with Off-Manifold Detergentcited by 71
  14. 2022Post hoc Explanations may be Ineffective for Detecting Unknown Spurious Correlationcited by 59
  15. 2020Evaluations and Methods for Explanation through Robustness Analysiscited by 51
  16. 2020SemEval-2020 Task 4: Commonsense Validation and Explanationcited by 42
  17. 2022A Comparative Study of Faithfulness Metrics for Model Interpretability Methodscited by 38
  18. 2019Can I Trust the Explainer? Verifying Post-hoc Explanatory Methodscited by 37
  19. 2021Evaluating the Faithfulness of Importance Measures in NLP by Recursively Masking Allegedly Important Tokens and Retrainingcited by 31
  20. 2021On the Sensitivity and Stability of Model Interpretations in NLPcited by 30
  21. 2022Explaining Chest X-ray Pathologies in Natural Languagecited by 27
  22. 2021Knowledge-Grounded Self-Rationalization via Extractive and Natural Language Explanationscited by 24
  23. 2021Generating Fluent Fact Checking Explanations with Unsupervised Post-Editingcited by 15
  24. 2020The Struggles of Feature-Based Explanations: Shapley Values vs. Minimal Sufficient Subsetscited by 14

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2023: 9; 2024: 30; 2025: 34; 2026: 34.

107 papers.

Sort:

By area:

  1. 2023Explainability for Large Language Models: A Surveycited by 331
  2. 2025Chain-of-Thought Reasoning In The Wild Is Not Always Faithfulcited by 166
  3. 2024Rethinking Interpretability in the Era of Large Language Modelscited by 110
  4. 2024Making Reasoning Matter: Measuring and Improving Faithfulness of Chain-of-Thought Reasoningcited by 108
  5. 2024Are self-explanations from Large Language Models faithful?cited by 101
  6. 2024FaithEval: Can Your Language Model Stay Faithful to Context, Even If "The Moon is Made of Marshmallows"cited by 85
  7. 2024RAVEL: Evaluating Interpretability Methods on Disentangling Language Model Representationscited by 77
  8. 2025Are DeepSeek R1 And Other Reasoning Models More Faithful?cited by 69
  9. 2025Thought Crime: Backdoors and Emergent Misalignment in Reasoning Modelscited by 57
  10. 2025MIB: A Mechanistic Interpretability Benchmarkcited by 49
  11. 2023Rigorously Assessing Natural Language Explanations of Neuronscited by 44
  12. 2025Measuring Chain of Thought Faithfulness by Unlearning Reasoning Stepscited by 43
38 more
  1. 2025Fostering Appropriate Reliance on Large Language Models: The Role of Explanations, Sources, and Inconsistenciescited by 39
  2. 2024Verification and Refinement of Natural Language Explanations through LLM-Symbolic Theorem Provingcited by 35
  3. 2023Quantifying Uncertainty in Natural Language Explanations of Large Language Modelscited by 31
  4. 2024The Probabilities Also Matter: A More Faithful Metric for Faithfulness of Free-Text Explanations in Large Language Modelscited by 29
  5. 2024How Interpretable are Reasoning Explanations from Prompting Large Language Models?cited by 27
  6. 2023People Make Better Edits: Measuring the Efficacy of LLM-Generated Counterfactually Augmented Data for Harmful Language Detectioncited by 25
  7. 2024Model Internals-based Answer Attribution for Trustworthy Retrieval-Augmented Generationcited by 25
  8. 2024Unsupervised Evaluation of Code LLMs with Round-Trip Correctnesscited by 23
  9. 2024Do Vision & Language Decoders use Images and Text equally? How Self-consistent are their Explanations?cited by 23
  10. 2025Self-Interpretability: LLMs Can Describe Complex Internal Processes that Drive Their Decisionscited by 23
  11. 2025Measuring the Faithfulness of Thinking Drafts in Large Reasoning Modelscited by 22
  12. 2024FinDVer: Explainable Claim Verification over Long and Hybrid-Content Financial Documentscited by 19
  13. 2025How does Chain of Thought Think? Mechanistic Interpretability of Chain-of-Thought Reasoning with Sparse Autoencodingcited by 19
  14. 2024xTower: A Multilingual LLM for Explaining and Correcting Translation Errorscited by 17
  15. 2025Explainable Chain-of-Thought Reasoning: An Empirical Analysis on State-Aware Reasoning Dynamicscited by 16
  16. 2025Measuring Chain-of-Thought Monitorability Through Faithfulness and Verbositycited by 15
  17. 2026Counterfactual Simulation Training for Chain-of-Thought Faithfulnesscited by 15
  18. 2024Towards Faithful Natural Language Explanations: A Study Using Activation Patching in Large Language Modelscited by 14
  19. 2025Support Evaluation for the TREC 2024 RAG Track: Comparing Human versus LLM Judgescited by 14
  20. 2024CausalEval: Towards Better Causal Reasoning in Language Modelscited by 13
  21. 2025A Causal Lens for Evaluating Faithfulness Metricscited by 12
  22. 2025A Closer Look at Bias and Chain-of-Thought Faithfulness of Large (Vision) Language Modelscited by 12
  23. 2026A Positive Case for Faithfulness: LLM Self-Explanations Help Predict Model Behaviorcited by 12
  24. 2024Multi-Level Explanations for Generative Language Modelscited by 11
  25. 2025Truthful or Fabricated? Using Causal Attribution to Mitigate Reward Hacking in Explanationscited by 9
  26. 2025Self-Critique and Refinement for Faithful Natural Language Explanationscited by 9
  27. 2023QA-NatVer: Question Answering for Natural Logic-based Fact Verificationcited by 8
  28. 2024Self-AMPLIFY: Improving Small Language Models with Self Post Hoc Explanationscited by 8
  29. 2024Cross-Refine: Improving Natural Language Explanation Generation by Learning in Tandemcited by 8
  30. 2025Do Activation Verbalization Methods Convey Privileged Information?cited by 8
  31. 2026Biases in the Blind Spot: Detecting What LLMs Fail to Mentioncited by 8
  32. 2024Towards a Framework for Evaluating Explanations in Automated Fact Verificationcited by 7
  33. 2024Faithful and Plausible Natural Language Explanations for Image Classification: A Pipeline Approachcited by 7
  34. 2026Do LLM Self-Explanations Help Users Predict Model Behavior? Evaluating Counterfactual Simulatability with Pragmatic Perturbationscited by 7
  35. 2025From Nodes to Narratives: Explaining Graph Neural Networks with LLMs and Graph Contextcited by 6
  36. 2025Beyond Correctness: Rewarding Faithful Reasoning in Retrieval-Augmented Generationcited by 6
  37. 2024Inference to the Best Explanation in Large Language Modelscited by 5
  38. 2024Fool Me Once? Contrasting Textual and Visual Explanations in a Clinical Decision-Support Settingcited by 5

50 most cited of 107 All 107 (JSON)