mlsanityData to 14 Oct 2026

Circuit Insights: Towards Interpretability Beyond Activations

arXiv 2510.14936 ↗First posted 16 Oct 2025cs.LG cs.AI cs.CLcited by 5

On the map: Mechanistic interpretability → Circuits and mechanistic analysis

Abstract (arXiv)

The fields of explainable AI and mechanistic interpretability aim to uncover the internal structure of neural networks, with circuit discovery as a central tool for understanding model computations. Existing approaches, however, rely on manual inspection and remain limited to toy tasks. Automated interpretability offers scalability by analyzing isolated features and their activations, but it often misses interactions between features and depends strongly on external LLMs and dataset quality. Transcoders have recently made it possible to separate feature attributions into input-dependent and input-invariant components, providing a foundation for more systematic circuit analysis. Building on this, we propose WeightLens and CircuitLens, two complementary methods that go beyond activation-based analysis. WeightLens interprets features directly from their learned weights, removing the need for explainer models or datasets while matching or exceeding the performance of existing methods on context-independent features. CircuitLens captures how feature activations arise from interactions between components, revealing circuit-level dynamics that activation-only approaches cannot identify. Together, these methods increase interpretability robustness and enhance scalable mechanistic analysis of circuits while maintaining efficiency and quality.

Key facts. Cited by 5 papers in the map (data through 14 Oct 2026). Placed in Circuits and mechanistic analysis, one of 13 directions in Mechanistic interpretability. Builds on 16 of its 29 references that are in the map. First posted on arXiv 16 Oct 2025.

Builds on

16 of 29 references.

Closest ancestors · 4

Same direction, newest first

  1. 2024Transcoders Find Interpretable LLM Feature Circuits17 Jun
  2. 2024Automatically Identifying Local and Global Circuits with Linear Computation Graphs22 May
  3. 2023Towards Automated Circuit Discovery for Mechanistic Interpretability28 Apr
  4. 2022Interpretability in the Wild: a Circuit for Indirect Object Identification in GPT-2 small1 Nov

Other references · 12

  1. 2020The Pile: An 800GB Dataset of Diverse Text for Language Modelingcited by 2,411
  2. 2022Large Language Models Encode Clinical Knowledgecited by 1,830
  3. 2023Towards General Text Embeddings with Multi-stage Contrastive Learningcited by 723
  4. 2024Gemma Scope: Open Sparse Autoencoders Everywhere All At Once on Gemma 2cited by 394
  5. 2019Unmasking Clever Hans Predictors and Assessing What Machines Really Learncited by 365
  6. 2024AttnLRP: Attention-Aware Layer-Wise Relevance Propagation for Transformerscited by 120
  7. 2024Automatically Interpreting Millions of Features in Large Language Modelscited by 113
  8. 2025Enhancing Automated Interpretability with Output-Centric Feature Descriptionscited by 41
  9. 2025Mechanistic understanding and validation of large AI models with SemanticLenscited by 29
  10. 2025FADE: Why Bad Descriptions Happen to Good Featurescited by 13
  11. 2025Capturing Polysemanticity with PRISM: A Multi-Concept Feature Description Frameworkcited by 10
  12. 2025Deceptive Automated Interpretability: Language Models Coordinating to Fool Oversight Systemscited by 3

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 1; 2026: 4.

5 papers.

  1. 2026Beyond Importance: Interchange-Sobol Sensitivity Reveals Task-Specific Content Channels in Transformer Componentscited by 3
  2. 2025Explainability of Large Language Models: Opportunities and Challenges toward Generating Trustworthy Explanationscited by 2
  3. 2026Judge Circuits Explain Format-Induced Inconsistency in LLM-as-a-Judgecited by 1
  4. 2026Mechanistic Interpretability for Neural Networks: Circuits, Sparse Features and Symbolic Reasoningcited by 1
  5. 2026DV-Lens: Revealing the Functional Organization of Language Model Parameterscited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2026Rethinking Circuit Evaluation: Do Circuits Explain Model Errors?
  2. 2024The Computational Complexity of Circuit Discovery for Inner Interpretability
  3. 2024Evaluating Brain-Inspired Modular Training in Automated Circuit Discovery for Mechanistic Interpretability
  4. 2026From Circuit Evidence to Mechanistic Theory: An Inductive Logic Approach
  5. 2026Formal Mechanistic Interpretability: Automated Circuit Discovery with Provable Guarantees
  6. 2026CircuitKIT : Circuit Discovery, Evaluation, and Application Toolkit for Mechanistic Interpretability
  7. 2026Seeing Through Circuits: Faithful Mechanistic Interpretability for Vision Transformers
  8. 2024Efficient Automated Circuit Discovery in Transformers using Contextual Decomposition