mlsanityData to 14 Oct 2026

Mechanistic Interpretability Needs Philosophy

arXiv 2506.18852 ↗First posted 23 Jun 2025cs.CL cs.AIcited by 6

On the map: Mechanistic interpretability → Circuits and mechanistic analysis

Abstract (arXiv)

Mechanistic interpretability (MI) aims to explain how neural networks work by uncovering their underlying mechanisms. As the field grows in influence, it is increasingly important to examine not just models themselves, but the assumptions, concepts and explanatory strategies implicit in MI research. We argue that mechanistic interpretability needs philosophy as an ongoing partner in clarifying its concepts, refining its methods, and navigating the epistemic and ethical complexities of interpreting AI systems. There is significant unrealised potential for progress in MI to be gained through deeper engagement with philosophers and philosophical frameworks. Taking three open problems from the MI literature as examples, this paper illustrates the value philosophy can add to MI research, and outlines a path toward deeper interdisciplinary dialogue.

Key facts. Cited by 6 papers in the map (data through 14 Oct 2026). Placed in Circuits and mechanistic analysis, one of 13 directions in Mechanistic interpretability. Builds on 53 of its 152 references that are in the map. First posted on arXiv 23 Jun 2025.

Builds on

53 of 152 references.

Closest ancestors · 8

Same direction, newest first

  1. 2025Everything, Everywhere, All at Once: Is Mechanistic Interpretability Identifiable?28 Feb
  2. 2025Open Problems in Mechanistic Interpretability27 Jan
  3. 2024Mechanistic?7 Oct
  4. 2024The Quest for the Right Mediator: Surveying Mechanistic Interpretability Through the Lens of Causal Mediation Analysis2 Aug
  5. 2024Mechanistic Interpretability for AI Safety -- A Review22 Apr
  6. 2023Is This the Subspace You Are Looking for? An Interpretability Illusion for Subspace Activation Patching28 Nov
  7. 2023Towards Automated Circuit Discovery for Mechanistic Interpretability28 Apr
  8. 2023Causal Abstraction: A Theoretical Foundation for Mechanistic Interpretability11 Jan

Other references · 45

  1. 2013Visualizing and Understanding Convolutional Networkscited by 2,290
  2. 2023Sparse Autoencoders Find Highly Interpretable Features in Language Modelscited by 1,420
  3. 2017Explanation in Artificial Intelligence: Insights from the Social Sciencescited by 1,206
  4. 2018A Survey Of Methods For Explaining Black Box Modelscited by 973
  5. 2022Toy Models of Superpositioncited by 971
  6. 2022Discovering Latent Knowledge in Language Models Without Supervisioncited by 767
  7. 2023The Linear Representation Hypothesis and the Geometry of Large Language Modelscited by 716
  8. 2023The Geometry of Truth: Emergent Linear Structure in Large Language Model Representations of True/False Datasetscited by 712
  9. 2023The Internal State of an LLM Knows When It's Lyingcited by 659
  10. 2020Null It Out: Guarding Protected Attributes by Iterative Nullspace Projectioncited by 455
  11. 2021Causal Abstractions of Neural Networkscited by 422
  12. 2023Function Vectors in Large Language Modelscited by 308
33 more
  1. 2023Large Language Models Fail on Trivial Alterations to Theory-of-Mind Taskscited by 292
  2. 2013Direct and Indirect Effectscited by 280
  3. 2019Risks from Learned Optimization in Advanced Machine Learning Systemscited by 258
  4. 2023Finding Alignments Between Interpretable Causal Variables and Distributed Neural Representationscited by 209
  5. 2023AI Deception: A Survey of Examples, Risks, and Potential Solutionscited by 188
  6. 2023The Clock and the Pizza: Two Stories in Mechanistic Explanation of Neural Networkscited by 156
  7. 2022Toward Transparent AI: A Survey on Interpreting the Inner Structures of Deep Neural Networkscited by 129
  8. 2021Inducing Causal Structure for Interpretable Neural Networkscited by 110
  9. 2020Explainable Automated Fact-Checking: A Surveycited by 83
  10. 2023How to Catch an AI Liar: Lie Detection in Black-Box LLMs by Asking Unrelated Questionscited by 83
  11. 2023Still No Lie Detector for Language Models: Probing Empirical and Conceptual Roadblockscited by 78
  12. 2024Alice in Wonderland: Simple Tasks Showing Complete Reasoning Breakdown in State-Of-the-Art Large Language Modelscited by 68
  13. 2024Talking Heads: Understanding Inter-layer Communication in Transformer Language Modelscited by 41
  14. 2023The Vector Grounding Problemcited by 39
  15. 2023Causal Representation Learning Made Identifiable by Grouping of Observational Variablescited by 32
  16. 2024Evaluating the Robustness of Analogical Reasoning in Large Language Modelscited by 31
  17. 2024Standards for Belief Representations in LLMscited by 30
  18. 2024On Behalf of the Stakeholders: Trends in NLP Model Interpretability in the Era of LLMscited by 25
  19. 2024Missed Causes and Ambiguous Effects: Counterfactuals Pose Challenges for Interpreting Neural Networkscited by 23
  20. 2022Necessity and Sufficiency for Explaining Text Classifiers: A Case Study in Hate Speech Detectioncited by 18
  21. 2024A Philosophical Introduction to Language Models - Part II: The Way Forwardcited by 18
  22. 2023Operationalising Representation in Natural Language Processingcited by 14
  23. 2024Does ChatGPT Have a Mind?cited by 14
  24. 2025Propositional Interpretability in Artificial Intelligencecited by 12
  25. 2024Position: An Inner Interpretability Framework for AI Inspired by Lessons from Cognitive Neurosciencecited by 11
  26. 2024Deception and Manipulation in Generative AIcited by 8
  27. 2024Defining Knowledge: Bridging Epistemology and Large Language Modelscited by 7
  28. 2025What Do Large Language Models Know? Tacit Knowledge as a Potential Causal-Explanatory Structurecited by 6
  29. 2024On the attribution of confidence to large language modelscited by 5
  30. 2024Real Sparks of Artificial Intelligence and the Importance of Inner Interpretabilitycited by 4
  31. 2024A Mechanistic Explanatory Strategy for XAIcited by 3
  32. 2025Can structural correspondences ground real world representational content in Large Language Models?cited by 3
  33. 2024Chatting with Bots: AI, Speech Acts, and the Edge of Assertioncited by 2

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 1; 2026: 5.

6 papers.

  1. 2025Interpretability as Alignment: Making Internal Understanding a Design Principlecited by 3
  2. 2026Mechanistic Interpretability of Brain-to-Speech Models Across Speech Modescited by 1
  3. 2026Vision-Language Asymmetry in Bistable Image Captioningcited by 0
  4. 2026Emergent Alignmentcited by 0
  5. 2026The Role of Rigor in Artificial Intelligencecited by 0
  6. 2026Deep Learning Models Also Recall Featurescited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025Unboxing the Black Box: Mechanistic Interpretability for Algorithmic Understanding of Neural Networks
  2. 2025Evaluating Explanations: An Explanatory Virtues Framework for Mechanistic Interpretability -- The Strange Science Part I.ii
  3. 2025A Mathematical Philosophy of Explanations in Mechanistic Interpretability -- The Strange Science Part I.i
  4. 2026Mechanistic Interpretability for Neural Networks: Circuits, Sparse Features and Symbolic Reasoning
  5. 2024Challenges in Mechanistically Interpreting Model Representations
  6. 2024Validating Mechanistic Interpretations: An Axiomatic Approach
  7. 2024The Cognitive Revolution in Interpretability: From Explaining Behavior to Interpreting Representations and Algorithms
  8. 2026Tracking Equivalent Mechanistic Interpretations Across Neural Networks