mlsanityData to 14 Oct 2026

States of LLM-generated Texts and Phase Transitions between them

arXiv 2503.06330 ↗First posted 8 Mar 2025cs.CL cond-mat.stat-mech cs.AIcited by 3

Not on the map. Why?

Abstract (arXiv)

It is known for some time that autocorrelations of words in human-written texts decay according to a power law. Recent works have also shown that the autocorrelations decay in texts generated by LLMs is qualitatively different from the literary texts. Solid state physics tie the autocorrelations decay laws to the states of matter. In this work, we empirically demonstrate that, depending on the temperature parameter, LLMs can generate text that can be classified as solid, critical state or gas.

Key facts. Cited by 3 papers in the map (data through 14 Oct 2026). Builds on 11 of its 37 references that are in the map. First posted on arXiv 8 Mar 2025.

Builds on

11 of 37 references.

References · 11

  1. 2019The Curious Case of Neural Text Degenerationcited by 2,916
  2. 2018Hierarchical Neural Story Generationcited by 1,193
  3. 2018Learning to Write with Cooperative Discriminatorscited by 119
  4. 2018Importance of Search and Evaluation Strategies in Neural Dialogue Modelingcited by 44
  5. 2016Criticality in Formal Languages and Statistical Physicscited by 35
  6. 2019Evaluating Computational Language Models with Scaling Properties of Natural Languagecited by 22
  7. 2017Do Neural Nets Learn Statistical Laws behind Natural Language?cited by 16
  8. 2019Mutual Information Scaling and Expressive Power of Sequence Modelscited by 14
  9. 2024Phase transition in large language models and the criticality of natural languagescited by 13
  10. 2023Autocorrelations Decay in Texts and Applicability Limits of Language Modelscited by 9
  11. 2018Natural Language Statistical Features of LSTM-generated Textscited by 7

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 3.

3 papers.

  1. 2026Temperature-driven inversion and nonlinear dynamics in ChatGPT-like AIscited by 1
  2. 2026Estimating Text Temperature with Language Modelscited by 0
  3. 2026Writing as a Self-Organized Critical Processcited by 0