Physics of generative AI’s atom: Repetition, bias, and beyond
Open Access
- 1 March 2026
- journal article
- Published by AIP Publishing in AIP Advances
- Vol. 16 (3)
Abstract
We derive a first-principles physics theory of the individual “atom” at the heart of generative AI such as ChatGPT: the basic attention head. The theory shows how, when, and why its output can become repetitive or suddenly switch to potentially harmful content. In situations where a small subset of attention heads dominates generative AI’s output or multi-layer effects average out, such undesirable microscale behavior will emerge at the macroscale to threaten generative AI’s safety in medical, legal, and business settings. The theory also quantifies the impact of bias from training and fine-tuning. The theory’s 2-spin form suggests why generative AI such as ChatGPT can work so well but hints that a generalized 3-spin attention might be even better. The theory’s similarity to spin bath physics means existing physics expertise could be harnessed to help generative AI become more trustworthy and resilient to manipulation.Keywords
This publication has 7 references indexed in Scilit:
- Training large language models on narrow tasks can lead to broad misalignmentNature, 2026
- Basic Attention Head as a Building Block toward Understanding Transformer-based Generative AIAdvances in Artificial Intelligence and Machine Learning, 2025
- Multispecies Cohesion: Humans, Machinery, AI, and BeyondPhysical Review Letters, 2024
- Mapping of attention mechanisms to a generalized Potts modelPhysical Review Research, 2024
- Transformer Variational Wave Functions for Frustrated Quantum Spin SystemsPhysical Review Letters, 2023
- Attention in Natural Language ProcessingIEEE Transactions on Neural Networks and Learning Systems, 2020
- Analyzing Multi-Head Self-Attention: Specialized Heads Do the Heavy Lifting, the Rest Can Be Pruned2019