mlsanityData to 7 Oct 2026

Temperature-driven inversion and nonlinear dynamics in ChatGPT-like AIs

arXiv 2608.00939 ↗First posted 2 Aug 2026physics.soc-ph cond-mat.dis-nn cs.AIcited by 1

Not on the map. Why?

Abstract (arXiv)

Increasing the temperature of an ordinary many-state system increases access to a wider range of states and hence increases its entropy. We find the opposite in ChatGPT-like AIs, even though raising the decoder temperature likewise increases access to a wider range of states (next-token choices). Across 12,000 continuations from 11 AIs, autoregressive feedback drives the long-time output population through an entropy maximum and into population inversion. The transition features frozen states, cycles, intermittency and noise-induced ordering. We present evidence of a hidden coordinate that acts as the state variable of an effective nonlinear map. Its trajectory average strongly predicts output repetition in separate test trajectories. ChatGPT-like AIs therefore behave not as `stochastic parrots', but as a new class of controllable nonlinear physical systems whose internal dynamics can be measured and perturbed.

Key facts. Cited by 1 paper in the map (data through 7 Oct 2026). Builds on 26 of its 47 references that are in the map. First posted on arXiv 2 Aug 2026.

Builds on

26 of 47 references.

References · 26

  1. 2024The Llama 3 Herd of Modelscited by 14,192
  2. 2015Distilling the Knowledge in a Neural Networkcited by 10,076
  3. 2021On the Opportunities and Risks of Foundation Modelscited by 3,961
  4. 2019The Curious Case of Neural Text Degenerationcited by 2,881
  5. 2022Large Language Models Encode Clinical Knowledgecited by 1,807
  6. 2023Pythia: A Suite for Analyzing Large Language Models Across Training and Scalingcited by 1,734
  7. 2023Sparse Autoencoders Find Highly Interpretable Features in Language Modelscited by 1,359
  8. 2023Progress measures for grokking via mechanistic interpretabilitycited by 822
  9. 2021Attention is Not All You Need: Pure Attention Loses Rank Doubly Exponentially with Depthcited by 411
  10. 2023Can ChatGPT Forecast Stock Price Movements? Return Predictability and Large Language Modelscited by 125
  11. 2025HalluLens: LLM Hallucination Benchmarkcited by 88
  12. 2020A Theoretical Analysis of the Repetition Problem in Text Generationcited by 83
14 more
  1. 2026Verbalizable Representations Form a Global Workspace in Language Modelscited by 59
  2. 2025The Curse of Depth in Large Language Modelscited by 58
  3. 2025Large Language Models Hallucination: A Comprehensive Surveycited by 45
  4. 2024Repetition Neurons: How Do Language Models Produce Repetitions?cited by 22
  5. 2025Unveiling Attractor Cycles in Large Language Models: A Dynamical Systems View of Successive Paraphrasingcited by 14
  6. 2024Phase transition in large language models and the criticality of natural languagescited by 12
  7. 2025Phase Transitions in Large Language Models and the O(N) Modelcited by 11
  8. 2025Shadows in the Attention: Contextual Perturbation and Representation Drift in the Dynamics of Hallucination in LLMscited by 9
  9. 2025From Condensation to Rank Collapse: A Two-Stage Analysis of Transformer Training Dynamicscited by 9
  10. 2024Phase Transitions in the Output Distribution of Large Language Modelscited by 7
  11. 2015Networkcontrologycited by 3
  12. 2025States of LLM-generated Texts and Phase Transitions between themcited by 2
  13. 2026LLMs as High-Dimensional Nonlinear Autoregressive Models with Attention: Training, Alignment and Inferencecited by 2
  14. 2026Generative Criticality in Large Language Model Temperature Scalingcited by 1

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2026: 1.

1 paper.

  1. 2026Interaction Creates Dynamical AI Behavior Absent in Isolationcited by 0