mlsanityData to 14 Oct 2026

Interpretability Can Be Actionable

arXiv 2605.11161 ↗First posted 11 May 2026cs.LG cs.AIcited by 7

Not on the map. Why?

Abstract (arXiv)

Interpretability aims to explain the behavior of deep neural networks. Despite rapid growth, there is mounting concern that much of this work has not translated into practical impact, raising questions about its relevance and utility. This position paper argues that the central missing ingredient is not new methods, but evaluation criteria: interpretability should be evaluated by actionability--the extent to which insights enable concrete decisions and interventions beyond interpretability research itself. We define actionable interpretability along two dimensions--concreteness and validation--and analyze the barriers currently preventing real-world impact. To address these barriers, we identify five domains where interpretability offers unique leverage and present a framework for actionable interpretability with evaluation criteria aligned with practical outcomes. Our goal is not to downplay exploratory research, but to establish actionability as a core objective of interpretability research.

Key facts. Cited by 7 papers in the map (data through 14 Oct 2026). Builds on 84 of its 128 references that are in the map. First posted on arXiv 11 May 2026.

Builds on

84 of 128 references.

References · 84

  1. 2021LoRA: Low-Rank Adaptation of Large Language Modelscited by 14,380
  2. 2023Mamba: Linear-Time Sequence Modeling with Selective State Spacescited by 4,718
  3. 2022Locating and Editing Factual Associations in GPTcited by 2,691
  4. 2022Language Models (Mostly) Know What They Knowcited by 1,856
  5. 2017Understanding Black-box Predictions via Influence Functionscited by 1,741
  6. 2017Towards A Rigorous Science of Interpretable Machine Learningcited by 1,416
  7. 2020Transformer Feed-Forward Layers Are Key-Value Memoriescited by 1,336
  8. 2023Inference-Time Intervention: Eliciting Truthful Answers from a Language Modelcited by 1,302
  9. 2017Explanation in Artificial Intelligence: Insights from the Social Sciencescited by 1,206
  10. 2022Interpretability in the Wild: a Circuit for Indirect Object Identification in GPT-2 smallcited by 1,165
  11. 2019Machine Unlearningcited by 1,061
  12. 2020Concept Bottleneck Modelscited by 1,032
38 more
  1. 2022Mass-Editing Memory in a Transformercited by 1,001
  2. 2022In-context Learning and Induction Headscited by 964
  3. 2023Steering Language Models With Activation Engineeringcited by 922
  4. 2023FrugalGPT: How to Use Large Language Models While Reducing Cost and Improving Performancecited by 853
  5. 2023Towards Automated Circuit Discovery for Mechanistic Interpretabilitycited by 724
  6. 2022What learning algorithm is in-context learning? Investigations with linear modelscited by 684
  7. 2023The Internal State of an LLM Knows When It's Lyingcited by 659
  8. 2020Estimating Training Data Influence by Tracing Gradient Descentcited by 580
  9. 2020Null It Out: Guarding Protected Attributes by Iterative Nullspace Projectioncited by 455
  10. 2024Mechanistic Interpretability for AI Safety -- A Reviewcited by 422
  11. 2023Unified Concept Editing in Diffusion Modelscited by 388
  12. 2025Persona Vectors: Monitoring and Controlling Character Traits in Language Modelscited by 345
  13. 2017The (Un)reliability of saliency methodscited by 333
  14. 2018Actionable Recourse in Linear Classificationcited by 328
  15. 2020Neural Additive Models: Interpretable Machine Learning with Neural Netscited by 319
  16. 2018On the Robustness of Interpretability Methodscited by 314
  17. 2022Language in a Bottle: Language Model Guided Concept Bottlenecks for Interpretable Image Classificationcited by 312
  18. 2019On the (In)fidelity and Sensitivity for Explanationscited by 310
  19. 2022Post-hoc Concept Bottleneck Modelscited by 293
  20. 2023Faithful Chain-of-Thought Reasoningcited by 288
  21. 2023Label-Free Concept Bottleneck Modelscited by 285
  22. 2019Approximating CNNs with Bag-of-local-Features models works surprisingly well on ImageNetcited by 260
  23. 2020Algorithmic Recourse: from Counterfactual Explanations to Interventionscited by 250
  24. 2024LLMs Know More Than They Show: On the Intrinsic Representation of LLM Hallucinationscited by 244
  25. 2019FACE: Feasible and Actionable Counterfactual Explanationscited by 233
  26. 2025AxBench: Steering LLMs? Even Simple Baselines Outperform Sparse Autoencoderscited by 218
  27. 2024Foundational Challenges in Assuring Alignment and Safety of Large Language Modelscited by 199
  28. 2024ReFT: Representation Finetuning for Language Modelscited by 197
  29. 2020Explaining Black Box Predictions and Unveiling Data Artifacts through Influence Functionscited by 144
  30. 2019Towards Realistic Individual Recourse and Actionable Explanations in Black-Box Decision Making Systemscited by 131
  31. 2023Editing Implicit Assumptions in Text-to-Image Diffusion Modelscited by 130
  32. 2020Learning to Faithfully Rationalize by Constructioncited by 113
  33. 2023Attention Sorting Combats Recency Bias In Long Context Language Modelscited by 90
  34. 2024Defending Against Unforeseen Failure Modes with Latent Adversarial Trainingcited by 88
  35. 2020Aligning Faithful Interpretations with their Social Attributioncited by 79
  36. 2025Artificial Intelligence Index Report 2025cited by 64
  37. 2023Cognitive Dissonance: Why Do Language Model Outputs Disagree with Internal Representations of Truthfulness?cited by 62
  38. 2025MIB: A Mechanistic Interpretability Benchmarkcited by 56

50 most cited of 84 All 84 (JSON)

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 7.

7 papers.

  1. 2026Tri-Info: Generalizable, Interpretable Failure Prediction for VLA Models via Information Theorycited by 3
  2. 2026Mechanistic Tomography: Designed Measurement for Control-Oriented Interpretabilitycited by 2
  3. 2026The Standard Interpretable Model: A general theory of interpretable machine learning to deductively design interpretable methods using Lagrangian mechanicscited by 1
  4. 2026ObserverBench: Testing Mechanistic Estimates for Intervention and Controlcited by 1
  5. 2026xMIx: High-Performance Serving-Time Platform for Mechanistic Interpretability Appscited by 0
  6. 2026Discovery and Spatial Characterisation of Multiple Shortcut Groups for Auditing Vision Model Biascited by 0
  7. 2026Shifting Mechanisms: How Positional Encoding Choice Shapes In-Context Retrievalcited by 0