{"id":"arxiv/2607.28667","url":"https://arxiv.org/abs/2607.28667","tags":["cs.LG","cs.AI","math.DS"],"score":0,"title":"Guarantees on Dynamical System Distinguishability for LLM Token Generation","domain":"cs.LG","source":"arxiv","authors":["Mohamed Akrout","Dan Wilson"],"reviews":0,"subject":"cs.LG","summary":"Recent work has shown that classifying large language models (LLMs)' responses can be distinguished by modeling token embeddings as trajectories of a black-box dynamical system (DS) and comparing prediction residuals of two DSs. Despite the empirical success of this dynamical approach, a theoretical understanding of why it works, how well it scales as a function of the token sequence, and when it transfers across embedding models remains lacking. We address these questions by formalizing the classification task as a binary hypothesis test between two stochastic linear DSs. We show that the…","category":"research","comments":0,"disputed":false,"division":"cs","mentions":0,"discussed":0,"published":"2026-07-25","reactions":{}}