Mechanistic Interpretability

Follow this topic Ask about this topic

The field seeks to explain how a model’s internal computations produce particular outputs, rather than relying only on input-output behavior. Researchers use sparse autoencoders and transcoders to identify features, and probes, attribution graphs, activation patching, ablation, and steering to trace or test their roles. Studies examine language-model tasks such as factual recall, reasoning, and refusal, as well as vision, audio, and scientific models. A central challenge is that a signal can be easy to decode without being causally used, while circuit explanations and interventions can vary across prompts, models, and analysis choices.

Latest papers

391 papers63 papers in the last four weeks, up 75% on the four weeks before. 0.4% of all new papers.

Deep Learning Models Also Recall Features
Aug 21, 2026Pierre BeckmannNeural Network Interpretability Mechanistic Interpretability
Ecole Polytechnique Fédérale de Lausanne (EPFL) · Idiap Research Institute
Feature Evolution and Migration during Vision Transformer Training
Aug 20, 2026Joonas Järve, Halil Ibrahim Aysel, Tarun Khajuria +1Neural Network Training Dynamics Vision Transformer
Institute of Computer Science University of Tartu Tartu, Estonia
Mechanistic Tomography: Designed Measurement for Control-Oriented Interpretability
Aug 19, 2026Vijay ErramilliMechanistic Interpretability Transformer Interpretability
Certified Mechanistic Interpretability: Lifting Single-Input Findings to Bounded Neighbourhoods
Aug 17, 2026Zhen Zhang, Yanliang Huang, Peng Xie +2Mechanistic Interpretability Neural Network Verification
School of Computation, Information and Technology Technical University of Munich, Germany
Emotion Across Speech and Faces: Shared Affective Mechanisms in Multimodal Foundation Models
Aug 17, 2026Xiutian Zhao, Luqi Sun, Björn Schuller +1Multimodal Foundation Models Multimodal Emotion Recognition
Center for Language and Speech Processing (CLSP), Johns Hopkins University, USA · Group on Language, Audio & Music (GLAM), Imperial College London, UK
SCOUT: Semantic Concept Discovery for Open-Vocabulary Editing of face Recognition Templates
Aug 17, 2026Leon Todorov, Peter Rot, Peter Peer +2Face Recognition Neural Network Interpretability
University of Ljubljana: 1Faculty of Electrical Engineering · Faculty of Computer and Information Science
Broken Symmetry in LLM Refusal: Answer Release Is More Local Than Refusal Restoration
Aug 16, 2026Yiqi Liu, Yang Wang, Songxin Wang +2LLM Refusal Behavior Refusal Mechanisms in Language Models
University of Manchester · Shanghai University of Finance and Economics
SpIn-ViT: Designing a Sparsity-Induced Vision Transformer That Is Mechanistically Interpretable
Aug 14, 2026Philip H. Lee, Parth PadalkarExplainable Image Classification Vision Transformer
Independent Researcher · Texas State University
SPICE: Simple Polysemantic Feature Interpretation via Clustering-based Explanation
Aug 14, 2026Sehyun Lee, Dahee Kwon, Damin Lee +1Neural Network Interpretability Mechanistic Interpretability
Korea Advanced Institute of Science and Technology (KAIST), Daejeon, Republic of Korea · INEEJI, Seongnam, Republic of Korea
SPEAR: Structure Property Explainability with Attention Regularization
Aug 13, 2026Aditya Raghavan, Utkarsh Pratiush, Dalton A. Pearl +4Mechanistic Interpretability Materials Property Prediction
University of Tennessee, Knoxville, TN, USA
Explanation Multiplicity: Circuit-Level Interpretability Evidence Does Not Survive Defensible Analytic Variation
Aug 13, 2026Ajay Pravin MahaleMechanistic Interpretability Explainability Evaluation
Hochschule Trier
Predictive Memory Localization: Forecasting Selective Intervention Paths from Internal Signals
Aug 13, 2026Jinhao Jing, Tian Zeyu, Lucas Qingyang Fang +4Selective Prediction Mechanistic Interpretability
The Chinese University of Hong Kong, Shenzhen · Shanghai Jiao Tong University · University of California, Santa Cruz +3
Mechanist: AI as a Scientific Instrument for Discovering the Mechanisms of Intelligence
Aug 12, 2026Mengru Wang, Junfeng Fang, Shuofei Qiao +17AI Agents for Scientific Discovery Mechanistic Interpretability
Zhejiang University · National University of Singapore · Heriot-Watt University
Measuring Semantic Abstractness of SAE Features via Nonlocality
Aug 11, 2026Chuqiao Lin, Shivaji Sondhi, Xiao-Liang QiSparse Autoencoders Mechanistic Interpretability
Rudolf Peierls Centre for Theoretical Physics, University of Oxford · Leinweber Institute for Theoretical Physics, Stanford University
Interpreting Language Model Hidden States at Scale
Aug 10, 2026Jordan Pettyjohn, Mansi Sakarvadia, Nathaniel Hudson +3LLM Interpretability Mechanistic Interpretability
University of Chicago · 2Illinois Institute of Technology · 3Argonne National Laboratory +1
Intrinsic Structure: Spectral Identifiability for Mechanistic Interpretability
Aug 10, 2026Ashim Dhor, Pin-Yu ChenRepresentation Identifiability Parameter Identifiability
1IISER Bhopal · 2IBM Research
Label-Free Parkinson's Disease Screening from Face and Voice through Mechanistic Interpretability
Aug 10, 2026Jiaheng Su, Yu SunMechanistic Interpretability Multimodal Anomaly Detection
Independent Researcher, Los Angeles, CA, USA · California State Polytechnic University, Pomona, CA, USA
Multilingual Emotion Neurons in Large Audio-Language Models
Aug 9, 2026Xiutian Zhao, Philipp Koehn, Björn Schuller +1Cross-Lingual Representation Alignment Neuron-Level Interpretability
Center for Language and Speech Processing (CLSP), Johns Hopkins University, USA · Group on Language, Audio & Music (GLAM), Imperial College London, UK
Can Graph Learning Learn Circuits?
Aug 9, 2026Chester Tan, Moritz Lampert, Courtney Maynard +3Graph Neural Networks Mechanistic Interpretability
Chair of Machine Learning for Complex Networks Center for AI and Data Science (CAIDAS) Julius-Maximilians-Universität Würzburg · Khoury College of Computer Sciences Northeastern University · Network Science Institute Northeastern University +1
Diffusion LLMs as Targets and Adversaries: Mechanistic Safety Exploits
Aug 7, 2026Elena Dumitrescu, Gert Lek, Lydia Y. Chen +1Adversarial Attacks on LLMs LLM Safety Alignment
Delft University of Technology, Delft, The Netherlands · University of Neuchâtel, Neuchâtel, Switzerland.
Finding Usable Weight Mechanisms with Tiled SVD
Aug 7, 2026Ash Manvi, Samreena TajreenLLM Interpretability Mechanistic Interpretability
Aquin Labs
MI-MIDI: Mechanistic Interpretability of Text-to-MIDI Generation Models via Probing, Lenses and Steering
Aug 6, 2026Jakub Poćwiardowski, Mateusz ModrzejewskiControllable Music Generation Symbolic Music Generation
Institute of Computer Science Warsaw University of Technology Warsaw, Poland
Visualizing Graph-to-Answer Mechanism Recovery in Materials-Science Hypothesis Generation
Aug 4, 2026Shashwat Sourav, Subhadeep Pal, Markus J. Buehler +4LLM Reasoning with Graphs Scientific Hypothesis Generation
Department of Physics, Washington University in St. Louis · Oak Ridge National Laboratory · Lawrence Berkeley National Laboratory +6
Sparse Weight Decomposition for Efficient Circuit Extraction
Aug 4, 2026Chuanhao Yan, Xuhan Huang, Yawen Duan +4Transformer Interpretability Mechanistic Interpretability
IQuest Research · Safe AI Forum · University of Oxford +2
V-FIND: Revealing the Intrinsic Forgery Knowledge Encoded in Video Forgery Detectors
Aug 4, 2026Shichao Kan, Chengpeng Hong, Jingtong Dou +8Explainable Deepfake Detection Mechanistic Interpretability
Central South University · The University of Sydney · NExT++ Research Centre, National University of Singapore
Mechanistic Interpretability-Guided Selective Fine-Tuning of Vision-Language Models for Centimeter-Level Flood Depth Estimation
Aug 2, 2026Nafis Fuad, Xiaodong Qian, Dongxiao ZhuVLM Adaptation Depth Estimation
Department of Civil & Environmental Engineering Wayne State University Detroit, MI, USA · Department of Civil & Environmental Engineering Institute for AI and Data Science (AIDaS) Wayne State University Detroit, MI, USA · Department of Computer Science, Institute for AI and Data Science (AIDaS) Wayne State University Detroit, MI, USA
Emergent Latent-State Computation under Stochastic Volatility
Jul 28, 2026Xiaoyu Huang, Lulu WangTransformer Interpretability Mechanistic Interpretability
Department of Mathematics Temple University Philadelphia, PA, USA · Department of Data Analytics Dickinson College Carlisle, PA, USA
From Semantics to Readout: Mechanistic Understanding of Audio Tokens after Fine-Tuning for Temporal Audio Grounding
Jul 28, 2026Yujian Ma, Jinqiu Sang, Ruizhe Li +2LLM Grounding LLM Fine-Tuning
1Shanghai Institute of Artificial Intelligence for Education, East China Normal University, China · School of Computer Science and Technology, East China Normal University, China · School of Computer Science, University of Birmingham, UK +1
CADENCE: A Cardiac Atom Dictionary for Interpretable Neural Concept Extraction from ECG Foundation Models
Jul 28, 2026Yixuan Duan, Arjun Naik, Sadeer Al-Kindi +1Electrocardiography Electrocardiogram Foundation Models
Department of Electrical and Computer Engineering Rice University Houston, Texas, USA · Department of Computer Science Rice University Houston, Texas, USA · Department of Cardiology Houston Methodist Hospital Houston, Texas, USA
Sparse Autoencoders Encode Both Concepts and Functions: The Downstream Geometry of Feature Effects
Jul 27, 2026Phu Gia Hoang, Anwoy Chatterjee, Tanmoy Chakraborty +2Sparse Autoencoders Mechanistic Interpretability
UKP Lab, Technical University of Darmstadt · Department of Electrical Engineering Indian Institute of Technology Delhi, India

Aug 21, 2026J/K move · Enter open · S save

Deep Learning Models Also Recall Features

Pierre Beckmann

Ecole Polytechnique Fédérale de Lausanne (EPFL) · Idiap Research Institute

Neural Network Interpretability Mechanistic Interpretability

Open paper AskPDF arXiv

Recent work in mechanistic interpretability has studied how large language models recall facts stored in their weights. This paper argues that factual recall points to something broader: a general kind of operation in deep learning models, which I call feature recall. The core observation is that a linear projection can be read as retrieving stored information scaled by input activations. I define feature recall, show it applies across architectures, and contrast it with the established paradigm of feature combination. I also consider how cases of feature recall might be mechanistically identified. The account gives philosophers a new conceptual tool for understanding deep learning, and points to empirical directions for mechanistic interpretability research.