Skip to main content
Cornell University
Learn about arXiv becoming an independent nonprofit.
We gratefully acknowledge support from
the Simons Foundation, Schmidt Sciences, Stockholm University, and all contributors.
Donate
arxiv logo > cs > arXiv:2602.02385

Help | Advanced Search

Computer Science > Machine Learning

(cs)
[Submitted on 2 Feb 2026]

Title:Transformers learn factored representations

Authors:Adam Shai, Loren Amdahl-Culleton, Casper L. Christensen, Henry R. Bigelow, Fernando E. Rosas, Alexander B. Boyd, Eric A. Alt, Kyle J. Ray, Paul M. Riechers
View a PDF of the paper titled Transformers learn factored representations, by Adam Shai and 8 other authors
View PDF HTML (experimental)
Abstract:Transformers pretrained via next token prediction learn to factor their world into parts, representing these factors in orthogonal subspaces of the residual stream. We formalize two representational hypotheses: (1) a representation in the product space of all factors, whose dimension grows exponentially with the number of parts, or (2) a factored representation in orthogonal subspaces, whose dimension grows linearly. The factored representation is lossless when factors are conditionally independent, but sacrifices predictive fidelity otherwise, creating a tradeoff between dimensional efficiency and accuracy. We derive precise predictions about the geometric structure of activations for each, including the number of subspaces, their dimensionality, and the arrangement of context embeddings within them. We test between these hypotheses on transformers trained on synthetic processes with known latent structure. Models learn factored representations when factors are conditionally independent, and continue to favor them early in training even when noise or hidden dependencies undermine conditional independence, reflecting an inductive bias toward factoring at the cost of fidelity. This provides a principled explanation for why transformers decompose the world into parts, and suggests that interpretable low dimensional structure may persist even in models trained on complex data.
Subjects: Machine Learning (cs.LG)
Cite as: arXiv:2602.02385 [cs.LG]
  (or arXiv:2602.02385v1 [cs.LG] for this version)
  https://doi.org/10.48550/arXiv.2602.02385
arXiv-issued DOI via DataCite

Submission history

From: Paul Riechers [view email]
[v1] Mon, 2 Feb 2026 17:49:06 UTC (12,467 KB)
Full-text links:

Access Paper:

    View a PDF of the paper titled Transformers learn factored representations, by Adam Shai and 8 other authors
  • View PDF
  • HTML (experimental)
  • TeX Source
view license
Current browse context:
cs.LG
< prev   |   next >
new | recent | 2026-02
Change to browse by:
cs

References & Citations

  • NASA ADS
  • Google Scholar
  • Semantic Scholar
export BibTeX citation Loading...

Bookmark

BibSonomy logo Reddit logo

arXivLabs: experimental projects with community collaborators

arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.

Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.

Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.

Which authors of this paper are endorsers? | Disable MathJax (What is MathJax?)
  • About
  • Help
  • contact arXivClick here to contact arXiv Contact
  • subscribe to arXiv mailingsClick here to subscribe Subscribe
  • Copyright
  • Privacy Policy
  • Web Accessibility Assistance
  • arXiv Operational Status