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.

How LLMs Are Persuaded: A Few Attention Heads, Rerouted
May 10, 2026Xiangkun Sun, Lingkai Kong, Aoqi Zhang +2Self-Attention LLM Interpretability
Northeastern University · Harvard University · Tsinghua University +1
Towards Effective Theory of LLMs: A Representation Learning Approach
May 10, 2026Muhammed Ustaomeroglu, Guannan QuLLM Interpretability Mechanistic Interpretability
Carnegie Mellon University Pittsburgh, PA 15213
Repeated-Token Counting Reveals a Dissociation Between Representations and Outputs
May 10, 2026Sohan VenkateshLLM Interpretability Mechanistic Interpretability
Manipal Institute of Technology Bengaluru
A Geometric Perspective on Next-Token Prediction in Large Language Models: Three Emerging Phases
May 9, 2026Gianfranco Lombardo, Giuseppe Trimigno, Stefano CagnoniNeural Representation Geometry LLM Interpretability
Department of Engineering and Architecture University of Parma Parma, IT 43125
From Mechanistic to Compositional Interpretability
May 9, 2026Ward Gauderis, Thomas Dooms, Steven T. Homer +2Mechanistic Interpretability Neural Network Interpretability
Vrije Universiteit Brussel, Brussels, Belgium · Independent · UK AI Security Institute, London, United Kingdom.
Architecture, Not Scale: Circuit Localization in Large Language Models
May 9, 2026Sohan VenkateshTransformer Interpretability Grouped-Query Attention
Manipal Institute of Technology Bengaluru.
Mechanistic Analysis of Alignment Algorithms in Language Models
May 9, 2026Aarush Sinha, Ishan Garg, Veeraraju Elluru +2LLM Interpretability Mechanistic Interpretability
University of Copenhagen · Independent · IIT Jodhpur +2
How Much Do Circuits Tell Us? Measuring the Consistency and Specificity of Language Model Circuits
May 8, 2026Michael Li, Nishant SubramaniCircuit Discovery Mechanistic Interpretability
Northeastern University · Carnegie Mellon University – Language Technologies Institute
Position: Mechanistic Interpretability Must Disclose Identification Assumptions for Causal Claims
May 8, 2026Zezheng Lin, Fengming LiuMechanistic Interpretability Causal Identifiability
Preprint.
Where's the Plan? Locating Latent Planning in Language Models with Lightweight Mechanistic Interventions
May 8, 2026Nicole Ma, Nick RuiLLM Planning Mechanistic Interpretability
Stanford University.
Tree SAE: Learning Hierarchical Feature Structures in Sparse Autoencoders
May 8, 2026Tue M. Cao, Hoang X. Nhat, Raed Alharbi +2Hierarchical Representation Learning Sparse Autoencoders
Hanoi University of Science and Technology, Hanoi, Vietnam · University of Florida, Florida, USA · Computer Science Department, Saudi Electronic University
What Cohort INRs Encode and Where to Freeze Them
May 8, 2026Vasiliki Sideri-Lampretsa, Sophie Starck, Robbie Holland +2Implicit Neural Representations Mechanistic Interpretability
Chair of AI in Healthcare and Medicine, Technical University of Munich, Germany · TUM University Hospital, Munich, Germany · Stanford Center for Artificial Intelligence in Medicine and Imaging, Stanford University, Stanford, CA, USA +3
PLOT: Progressive Localization via Optimal Transport in Neural Causal Abstraction
May 7, 2026Jonathn Chang, Arya Datla, Ziv GoldfeldMechanistic Interpretability Causal Abstraction
Cornell University
Beyond the Black Box: Interpretability of Agentic AI Tool Use
May 7, 2026Hariom Tatsat, Ariye ShaterAI Agent Reliability Mechanistic Interpretability
Quantitative Analytics, Barclays
SoftSAE: Dynamic Top-K Selection for Adaptive Sparse Autoencoders
May 7, 2026Jakub Stępień, Marcin Mazur, Jacek Tabor +1Sparse Autoencoders Mechanistic Interpretability
Jagiellonian University · IDEAS Research Institute
Patch-Effect Graph Kernels for LLM Interpretability
May 7, 2026Ruben Fernandez-Boullon, David N. OlivieriGraph Representation Learning Mechanistic Interpretability
Department of Computer Science, University of Vigo, Ourense 32004, Spain
SMolLM: Small Language Models Learn Small Molecular Grammar
May 7, 2026Akhil Jindal, Harang JuSmall Language Models Mechanistic Interpretability
Moku
Navigating by Old Maps: The Pitfalls of Static Mechanistic Localization in LLM Post-Training
May 7, 2026Hang Chen, Jiaying Zhu, Hongyang Chen +3Supervised Fine-Tuning Transformer Interpretability
School of Computer Science and Technology Xi’an Jiaotong University · School of Computer Science and Engineering The Chinese University of Hong Kong · Shaanxi Co., Ltd(Xi’an 710077, China) China Mobile Group +1
Feature Starvation as Geometric Instability in Sparse Autoencoders
May 6, 2026Faris Chaudhry, Keisuke Yano, Anthea MonodRepresentation Learning Sparse Autoencoders
Imperial College London South Kensington Campus, London SW7 2AZ, UK · Institute for Statistical Mathematics 10-3 Midori-cho, Tachikawa Tokyo 190-8562, Japan
Retrieval and competition: how a protein foundation model starts a protein
May 5, 2026Piotr Jedryszek, Oliver M. CrookProtein Language Models Mechanistic Interpretability
Department of Biology, University of Oxford, Oxford, UK · Kavli Institute for Nanoscience Discovery, University of Oxford, Oxford, UK · Department of Chemistry, University of Oxford, Oxford, UK
What Happens Inside Agent Memory? Circuit Analysis from Emergence to Diagnosis
May 5, 2026Xutao Mao, Jinman Zhao, Gerald Penn +1Agent Memory Mechanistic Interpretability
City University of Hong Kong · University of Toronto
Pairwise matrices for sparse autoencoders: single-feature inspection mislabels causal axes
May 4, 2026Michael A. Riegler, Birk Sebastian Frostelid Torpmann-HagenSparse Autoencoders Mechanistic Interpretability
SimulaMet, Norway · Simula, Norway
Neuron-Anchored Rule Extraction for Large Language Models via Contrastive Hierarchical Ablation
May 4, 2026Francesco Sovrano, Gabriele Dominici, Marc LangheinrichLLM Interpretability Mechanistic Interpretability
Università della Svizzera italiana Lugano, Switzerland
Bucketing the Good Apples: A Method for Diagnosing and Improving Causal Abstraction
May 4, 2026Li Puyin, Jiyuan Tan, Ahmad Jabbar +2Mechanistic Interpretability Causal Abstraction
♢Stanford University · ♠Goodfire
Automated Interpretability and Feature Discovery in Language Models with Agents
May 2, 2026Arnau Marin-Llobet, Javier FerrandoMechanistic Interpretability
Harvard University
Arithmetic in the Wild: Llama uses Base-10 Addition to Reason About Cyclic Concepts
May 1, 2026Sheridan Feucht, Tal Haklay, Usha Bhalla +9Numerical Reasoning in Language Models LLM Interpretability
aNortheastern University · bTechnion IIT · cHarvard University +1
From Local to Global to Mechanistic: An iERF-Centered Unified Framework for Interpreting Vision Models
May 1, 2026Yearim Kim, Sangyu Han, Nojun KwakSaliency Map Evaluation Mechanistic Interpretability
Seoul National University, Seoul 08826, South Korea
Perturbation Probing: A Two-Pass-per-Prompt Diagnostic for FFN Behavioral Circuits in Aligned LLMs
Apr 30, 2026Hongliang Liu, Tung-Ling Li, Yuhao WuLLM Interpretability Mechanistic Interpretability
Palo Alto Networks
Density Field State Space Models: 1-Bit Distillation, Efficient Inference, and Knowledge Organization in Mamba-2
Apr 28, 2026Chirag ShindeModel Compression LLM Quantization
Independent Researcher
Differentiable Faithfulness Alignment for Cross-Model Circuit Transfer
Apr 27, 2026Shun Shao, Binxu Wang, Shay B. Cohen +2LLM Interpretability Mechanistic Interpretability
University of Cambridge · Kempner Institute, Harvard University · University of Edinburgh +1

May 10, 2026J/K move · Enter open · S save

How LLMs Are Persuaded: A Few Attention Heads, Rerouted

Xiangkun Sun, Lingkai Kong, Aoqi Zhang, Liang Zeng+1

Northeastern University · Harvard University · Tsinghua University+1

Self-Attention LLM Interpretability Mechanistic Interpretability Persuasion in Language Models Language Model Robustness

Open paper AskPDF arXiv

Large language models can answer a factual question correctly, yet switch to an incorrect answer when exposed to persuasive context. Why this happens inside the model remains unclear. In this work, we study this question through causal interventions in controlled multiple-choice tasks and identify that persuasion mainly acts through a small set of mid-layer attention heads, which we call decision heads. These heads do not compute the answer themselves. Instead, they copy the option they attend to, so changing their attention can directly change the model's answer. We find that this attention is controlled by a one-dimensional option-routing feature in the option-token representations. Persuasive text changes this feature, causing the decision heads to attend to the persuasion target instead of the correct option. We further trace the routing feature to shallower attention heads that read persuasive keywords from the input. Together, these results reveal a causal pathway: shallow heads read persuasive keywords and write the routing feature, which redirects the decision heads' attention toward the persuasion target; the decision heads then copy the selected option, leading to the wrong answer. We verify each step through intervention and find the same mechanism across four open-source model families and in a source-selection task derived from Generative Engine Optimization (GEO). Finally, based on this mechanism, we introduce a low-rank update to the decision heads' key projections to reduce their sensitivity to persuasion.