Skip to content
DraconiticResearch
Search Sources API
201,818 studies31 sourcesUpdated 54 min ago
PreprintCC BY 4.0arXiv

Notes on algebraic perspective on neural network conformal theories

  • Sam Leutheusser
  • Joydeep Naskar
Open at sourcePDFDOICite

Abstract

Neural network conformal theories were first introduced in \cite{Halverson:2024axc} and exact two-, three- and four-point correlators were computed from an ensemble of infinite-width neural networks. In particular, the four-point correlator of any scalar operator satisfied crossing-symmetry and conformal block decomposition across different channels were in agreement without any constraint on moments. An infinite tower of degenerate operators with the same scaling dimension were found. In this note, we take a closer look at the structure of correlators and relation between operators. We also layout a recipe to construct the algebra of the theory.

Related

  1. Neural Ideals and Neural Codes: An Algebraic Framework for Neural Network Classification and Feature Interpretation
    Venkata Subbaiah Yerrapati, Rahul Dixit, Ajay Kumar Shukla
    arXivPreprintCC BY 4.0Full textAlgebra & number theoryarXiv
  2. A Story of Algebra and Categories I: Algebraic Theories
    Lior Yanovski
    arXivPreprintCC BY 4.0Full textAlgebra & number theoryGeometry & topologyarXiv
  3. "Massive" Conformal Field Theories and application to Nambu-Goto string
    Yuri Makeenko
    arXivPreprintarXiv licenceFull textMathematical physicsHigh-energy & particle physicsarXiv
  4. Algebraic overtwistedness of contact (+1)-surgeries via augmented Floer theories
    Russell Avdek, Youlin Li, Zhengyi Zhou
    arXivPreprintCC BY 4.0Full textGeometry & topologyarXiv
  5. Exactly solvable conformal field theories
    Sylvain Ribault
    SciPost Physics Lecture NotesPeer-reviewedCC BY17 citedGeometry & topologyOpenAlex

Cite

BibTeXRIS
BibTeX
@misc{leutheusser2026notes,
  title = {{Notes on algebraic perspective on neural network conformal theories}},
  author = {Sam Leutheusser and Joydeep Naskar},
  howpublished = {arXiv},
  year = {2026},
  month = sep,
  doi = {10.48550/arxiv.2609.37074},
  eprint = {2609.37074},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2609.37074},
  note = {CC BY 4.0},
}
RIS
TY  - UNPB
TI  - Notes on algebraic perspective on neural network conformal theories
AU  - Sam Leutheusser
AU  - Joydeep Naskar
PY  - 2026
DA  - 2026/09/29
PB  - arXiv
DO  - 10.48550/arxiv.2609.37074
UR  - https://arxiv.org/abs/2609.37074
AB  - Neural network conformal theories were first introduced in \cite{Halverson:2024axc} and exact two-, three- and four-point correlators were computed from an ensemble of infinite-width neural networks. In particular, the four-point correlator of any scalar operator satisfied crossing-symmetry and conformal block decomposition across different channels were in agreement without any constraint on moments. An infinite tower of degenerate operators with the same scaling dimension were found. In this note, we take a closer look at the structure of correlators and relation between operators. We also layout a recipe to construct the algebra of the theory.
ER  -