English
Related papers

Related papers: Towards Automated Circuit Discovery for Mechanisti…

200 papers

We introduce methods for discovering and applying sparse feature circuits. These are causally implicated subnetworks of human-interpretable features for explaining language model behaviors. Circuits identified in prior work consist of…

Machine Learning · Computer Science 2025-03-28 Samuel Marks , Can Rager , Eric J. Michaud , Yonatan Belinkov , David Bau , Aaron Mueller

As AI systems are used in high-stakes applications, ensuring interpretability is crucial. Mechanistic Interpretability (MI) aims to reverse-engineer neural networks by extracting human-understandable algorithms to explain their behavior.…

Machine Learning · Computer Science 2025-03-03 Maxime Méloux , Silviu Maniu , François Portet , Maxime Peyrard

Transformers generate valid and diverse chemical structures, but little is known about the mechanisms that enable these models to capture the rules of molecular representation. We present a mechanistic analysis of autoregressive…

Machine Learning · Computer Science 2025-12-11 Kristof Varadi , Mark Marosi , Peter Antal

Mechanistic Interpretability (MI) aims to reverse-engineer model behaviors by identifying functional sub-networks. Yet, the scientific validity of these findings depends on their stability. In this work, we argue that circuit discovery is…

Machine Learning · Computer Science 2026-02-04 Maxime Méloux , François Portet , Maxime Peyrard

Recent advances in language model interpretability have identified circuits, critical subnetworks that replicate model behaviors, yet how knowledge is structured within these crucial subnetworks remains opaque. To gain an understanding…

Computation and Language · Computer Science 2025-07-17 Huaizhi Ge , Frank Rudzicz , Zining Zhu

Mechanistic interpretability aims to understand how neural networks generalize beyond their training data by reverse-engineering their internal structures. We introduce patterning as the dual problem: given a desired form of generalization,…

Machine Learning · Computer Science 2026-01-21 George Wang , Daniel Murfet

Mechanistic interpretability aims to understand the internal mechanisms learned by neural networks. Despite recent progress toward this goal, it remains unclear how best to decompose neural network parameters into mechanistic components. We…

Machine Learning · Computer Science 2025-02-11 Dan Braun , Lucius Bushnaq , Stefan Heimersheim , Jake Mendel , Lee Sharkey

Understanding AI systems' inner workings is critical for ensuring value alignment and safety. This review explores mechanistic interpretability: reverse engineering the computational mechanisms and representations learned by neural networks…

Artificial Intelligence · Computer Science 2024-08-27 Leonard Bereska , Efstratios Gavves

End-to-end deep neural networks have achieved remarkable success across various domains but are often criticized for their lack of interpretability. While post hoc explanation methods attempt to address this issue, they often fail to…

Machine Learning · Computer Science 2025-01-22 Weixin Chen , Simon Yu , Huajie Shao , Lui Sha , Han Zhao

Following their success across many domains, transformers have also proven effective for symbolic regression (SR); however, the internal mechanisms underlying their generation of mathematical operators remain largely unexplored. Although…

Machine Learning · Computer Science 2026-02-04 Arco van Breda , Erman Acar

Understanding the internal circuits that language models use to solve tasks remains a central challenge in mechanistic interpretability. A crucial part of finding circuits is understanding why each attention head attends where it does. To…

Machine Learning · Computer Science 2026-05-15 Gabriel Franco , Lucas M. Tassis , Azalea Rohr , Mark Crovella

Attention mechanisms form a core component of several successful deep learning architectures, and are based on one key idea: ''The output depends only on a small (but unknown) segment of the input.'' In several practical applications like…

Machine Learning · Computer Science 2023-05-16 Lakshmi Narayan Pandey , Rahul Vashisht , Harish G. Ramaswamy

Mechanistic interpretability often identifies circuits inside Transformer models, but explanations of those circuits are usually validated through examples, ablations, and manual reasoning. This leaves a gap between finding a plausible…

Machine Learning · Computer Science 2026-05-26 Neel Somani

Which components in transformer language models are responsible for discourse understanding? We hypothesize that sparse computational graphs, termed as discursive circuits, control how models process discourse relations. Unlike simpler…

Computation and Language · Computer Science 2025-10-14 Yisong Miao , Min-Yen Kan

Mechanistic interpretability aims to explain neural model behaviour by reverse-engineering learned computational structure into human-understandable components. Without a formal framework, however, mechanistic explanations cannot be…

Machine Learning · Computer Science 2026-05-12 Ward Gauderis , Thomas Dooms , Steven T. Holmer , Kola Ayonrinde , Geraint A. Wiggins

One of the main challenges in mechanistic interpretability is circuit discovery, determining which parts of a model perform a given task. We build on the Mechanistic Interpretability Benchmark (MIB) and propose three key improvements to…

Computation and Language · Computer Science 2025-10-31 Yaniv Nikankin , Dana Arad , Itay Itzhak , Anja Reusch , Adi Simhi , Gal Kesten-Pomeranz , Yonatan Belinkov

Neural networks often exhibit emergent behavior, where qualitatively new capabilities arise from scaling up the amount of parameters, training data, or training steps. One approach to understanding emergence is to find continuous…

Machine Learning · Computer Science 2023-10-23 Neel Nanda , Lawrence Chan , Tom Lieberum , Jess Smith , Jacob Steinhardt

Mechanistic interpretability often uses activation patching, causal tracing, path patching, and steering directions to reveal behaviorally meaningful directions in Transformer activation space. This paper develops a field-theoretic…

Machine Learning · Computer Science 2026-05-26 David N. Olivieri , Antonio F. Pérez Rodríguez

Transformers demonstrate impressive performance on a range of reasoning benchmarks. To evaluate the degree to which these abilities are a result of actual reasoning, existing work has focused on developing sophisticated benchmarks for…

Machine Learning · Computer Science 2024-07-02 Jannik Brinkmann , Abhay Sheshadri , Victor Levoso , Paul Swoboda , Christian Bartelt

Interpretability provides a toolset for understanding how and why neural networks behave in certain ways. However, there is little unity in the field: most studies employ ad-hoc evaluations and do not share theoretical foundations, making…