The Geometry of Concepts: Sparse Autoencoder Feature Structure.

Yuxiao Li, Eric J Michaud, David D Baek, Joshua Engels, Xiaoqing Sun, Max Tegmark

Abstract read

In one paragraph

Article in Entropy (Basel, Switzerland), 2025. The graph could read no effect estimate from its abstract, so it casts no vote on the map. Cited by 2 papers.

0numbers the graph read from it

0cells of the map it votes in

2citing papers in PubMed

–field-weighted citation impact

1 · What the graph read from it

What it found

Each row is one number read from the abstract, on the scale the paper reported it, with its interval. Left of the dashed line favours the treatment, right favours the comparator. Under each row is the sentence it came from. New to these charts? A ten-minute tutorial.

The abstract states no effect estimate the extractor could read, or names no intervention and outcome on the map, so this paper lights no cell and moves no belief. It is still indexed, cited and linked below.

2 · The registry

The trial behind it

Trials whose registry record cites this paper, or whose number appears in the abstract. A trial that started after this paper was published is citing it as background, not reporting it.

Neither the registry nor the abstract names a trial number. If this is a trial report, that itself is worth knowing.

3 · Its place in the literature

Who cites it

2 citing papers in PubMed.

Attention heads of large language models.Patterns (New York, N.Y.) · 2025
Review
Spontaneous Emergence of Agent Individuality Through Social Interactions in Large Language Model-Based Communities.Entropy (Basel, Switzerland) · 2024
Article

4 · The record

Corrections and comments

PubMed lists nothing against this paper. Absence here is not a guarantee, only a check that was made.

5 · Who and what money

Authors and funding

6 authors.

Yuxiao LiBeneficial AI Foundation (BAIF), Cambridge, MA 02139, USA.ORCID 0000-0002-6496-9991

Eric J MichaudDepartment of Physics, Massachusetts Institute of Technology, Cambridge, MA 02139, USA.

David D BaekInstitute for Artificial Intelligence and Fundamental Interaction, Cambridge, MA 02139, USA.

Joshua EngelsDepartment of Electrical Engineering and Computer Science, Massachusetts Institute of Technology, Cambridge, MA 02139, USA.

Xiaoqing SunDepartment of Physics, Massachusetts Institute of Technology, Cambridge, MA 02139, USA.

Max TegmarkDepartment of Physics, Massachusetts Institute of Technology, Cambridge, MA 02139, USA.ORCID 0000-0001-7670-7190

Funding

U.S. National Science Foundation 2141064U.S. National Science Foundation PHY-2019786

6 · The paper itself

Abstract

Sparse autoencoders have recently produced dictionaries of high-dimensional vectors corresponding to the universe of concepts represented by large language models. We find that this concept universe has interesting structure at three levels: (1) The "atomic" small-scale structure contains "crystals" whose faces are parallelograms or trapezoids, generalizing well-known examples such as (

Indexed as

clusteringlarge language modelsmechanistic interpretabilityneural networkssparse coding

Identifiers

PMID40282579

DOI10.3390/e27040344

PMCPMC12025678

What OpenQuestion holds

Textmetadata

LicenceCC BY

Read underepoch 390

Registered trials

None linked

Read under generation 80e0d062 · epoch 390. Bibliography from PubMed, PubMed Central and OpenAlex; grants from NIH RePORTER; trial links from ClinicalTrials.gov; estimates, votes and beliefs from the OpenQuestion graph.