English
Related papers

Related papers: The Geometry of Concepts: Sparse Autoencoder Featu…

200 papers

A popular new method in mechanistic interpretability is to train high-dimensional sparse autoencoders (SAEs) on neuron activations and use SAE features as the atomic units of analysis. However, the body of evidence on whether SAE feature…

Machine Learning · Computer Science 2024-09-10 Maheep Chaudhary , Atticus Geiger

Sparse autoencoders (SAEs) are a recent technique for decomposing neural network activations into human-interpretable features. However, in order for SAEs to identify all features represented in frontier models, it will be necessary to…

Machine Learning · Computer Science 2025-06-04 Anish Mudide , Joshua Engels , Eric J. Michaud , Max Tegmark , Christian Schroeder de Witt

Sparse Autoencoders (SAEs) provide potentials for uncovering structured, human-interpretable representations in Large Language Models (LLMs), making them a crucial tool for transparent and controllable AI systems. We systematically analyze…

Machine Learning · Computer Science 2026-02-03 Jack Gallifant , Shan Chen , Kuleen Sasse , Hugo Aerts , Thomas Hartvigsen , Danielle S. Bitterman

As AI models achieve remarkable capabilities across diverse domains, understanding what representations they learn and how they encode concepts has become increasingly important for both scientific progress and trustworthy deployment.…

Machine Learning · Computer Science 2026-05-05 Yiming Tang , Harshvardhan Saini , Zhaoqian Yao , Zheng Lin , Yizhen Liao , Jingyi Cui , Yisen Wang , Mengnan Du , Dianbo Liu

Concept learning constructs visual representations that are connected to linguistic semantics, which is fundamental to vision-language tasks. Although promising progress has been made, existing concept learners are still vulnerable to…

Computer Vision and Pattern Recognition · Computer Science 2024-04-01 Qi Zheng , Chaoyue Wang , Dadong Wang , Dacheng Tao

Leveraging on the underlying low-dimensional structure of data, low-rank and sparse modeling approaches have achieved great success in a wide range of applications. However, in many applications the data can display structures beyond simply…

Machine Learning · Computer Science 2019-12-04 Zhao Kang , Xiao Lu , Yiwei Lu , Chong Peng , Zenglin Xu

We introduce a new approach to systematically map features discovered by sparse autoencoder across consecutive layers of large language models, extending earlier work that examined inter-layer feature links. By using a data-free cosine…

Machine Learning · Computer Science 2025-07-28 Daniil Laptev , Nikita Balagansky , Yaroslav Aksenov , Daniil Gavrilov

Category-selectivity in the brain describes the observation that certain spatially localized areas of the cerebral cortex tend to respond robustly and selectively to stimuli from specific limited categories. One of the most well known…

Neurons and Cognition · Quantitative Biology 2021-12-21 T. Anderson Keller , Qinghe Gao , Max Welling

Sparse Autoencoders (SAEs) have become an important tool in mechanistic interpretability, helping to analyze internal representations in both Large Language Models (LLMs) and Vision Transformers (ViTs). By decomposing polysemantic…

Machine Learning · Computer Science 2026-05-11 Jakub Stępień , Marcin Mazur , Jacek Tabor , Przemysław Spurek

Sparse autoencoders (SAEs) model the activations of a neural network as linear combinations of sparsely occurring directions of variation (latents). The ability of SAEs to reconstruct activations follows scaling laws w.r.t. the number of…

Machine Learning · Computer Science 2025-09-05 Eric J. Michaud , Liv Gorton , Tom McGrath

This paper seeks to combine dictionary learning and hierarchical image representation in a principled way. To make dictionary atoms capturing additional information from extended receptive fields and attain improved descriptive capacity, we…

Computer Vision and Pattern Recognition · Computer Science 2019-11-11 Tong Zhang , Fatih Porikli

The coexistence of sparsity and clustering (non-vanishing average fraction of triangles per node) is one of the few structural features that, irrespective of finer details, are ubiquitously observed across large real-world networks. This…

Probability · Mathematics 2026-03-17 Alessio Catanzaro , Remco van der Hofstad , Diego Garlaschelli

Autoencoders are a prominent model in many empirical branches of machine learning and lossy data compression. However, basic theoretical questions remain unanswered even in a shallow two-layer setting. In particular, to what degree does a…

Machine Learning · Computer Science 2024-02-08 Kevin Kögler , Alexander Shevchenko , Hamed Hassani , Marco Mondelli

Sparse autoencoders (SAEs) are a technique for sparse decomposition of neural network activations into human-interpretable features. However, current SAEs suffer from feature absorption, where specialized features capture instances of…

Machine Learning · Computer Science 2025-09-29 Anton Korznikov , Andrey Galichin , Alexey Dontsov , Oleg Rogov , Elena Tutubalina , Ivan Oseledets

While clusters of galaxies are considered one of the most important cosmological probes, the standard spherical modelling of the dark matter and the intracluster medium is only a rough approximation. Indeed, it is well established both…

Cosmology and Nongalactic Astrophysics · Physics 2015-06-11 Marceau Limousin , Andrea Morandi , Mauro Sereno , Massimo Meneghetti , Stefano Ettori , Matthias Bartelmann , Tomas Verdugo

Sparse Autoencoders (SAEs) have emerged as a powerful paradigm for disentangling feature superposition in transformer-based architectures, enabling precise control via activation steering. However, the theoretical foundations of…

Machine Learning · Computer Science 2026-05-08 Yunpeng Zhou

There is a large ongoing scientific effort in mechanistic interpretability to map embeddings and internal representations of AI systems into human-understandable concepts. A key element of this effort is the linear representation…

Machine Learning · Computer Science 2025-05-27 Alexander Modell , Patrick Rubin-Delanchy , Nick Whiteley

Structures in low-dimensional topology and low-dimensional geometry -- often combined with ideas from (quantum) field theory -- can explain and inspire concepts in algebra and in representation theory and their categorified versions. We…

Representation Theory · Mathematics 2015-11-09 Jürgen Fuchs , Christoph Schweigert

Black-box probing models can reliably extract linguistic features like tense, number, and syntactic role from pretrained word representations. However, the manner in which these features are encoded in representations remains poorly…

Computation and Language · Computer Science 2021-09-15 Evan Hernandez , Jacob Andreas

Large Language Models achieve next-token prediction by transporting a vectorized piece of text (prompt) across an accompanying embedding space under the action of successive transformer layers. The resulting high-dimensional trajectories…

Machine Learning · Computer Science 2025-02-17 Raphaël Sarfati , Toni J. B. Liu , Nicolas Boullé , Christopher J. Earls