---
title: "Stochastic Dynamics of Generative AI: From Autoregressive Models to Interacting LLM Networks and Neural SDEs"
source: "arXiv AI/CL/LG（论文）"
source_tier: "t2"
url: "https://arxiv.org/abs/2611.07065v1"
permalink: "https://feed.astralor.com/items/itm_20d485af3f89fdbfc57a8509"
published: "2026-11-04T19:23:13.000Z"
score: 57
channel: "tech"
tags: ["LLM","Stochastic Dynamics","Generative","From Autoregressive Models","Interacting LLM Networks","Neural SDEs This","First","LLMs"]
---

# Stochastic Dynamics of Generative AI: From Autoregressive Models to Interacting LLM Networks and Neural SDEs

## 速览

This paper develops a stochastic dynamical-systems perspective on three classes of generative AI models. First, we formulate large language models (LLMs) as high-dimensional nonlinear autoregressive models with attentio…

## 关键点

- This paper develops a stochastic dynamical-systems perspective on three classes of generative AI models.
- First, we formulate large language models (LLMs) as high-dimensional nonlinear autoregressive models with attention-based dependencies.
- This formulation connects pretraining by conditional maximum likelihood to sequential token generation and provides a probabilistic interpretation of LLM operation.

## 全文

This paper develops a stochastic dynamical-systems perspective on three classes of generative AI models. First, we formulate large language models (LLMs) as high-dimensional nonlinear autoregressive models with attention-based dependencies. This formulation connects pretraining by conditional maximum likelihood to sequential token generation and provides a probabilistic interpretation of LLM operation. Second, we consider networks of interacting LLMs and characterize their collective behavior through mean-field ordinary differential equations. These dynamics provide a framework for understanding emergent preferential attachment and persistent structural biases, including type-dependent glass-ceiling effects. Third, we consider neural stochastic differential equations (SDEs) as continuous-time generative models for stochastic trajectories, including those observed at irregular time intervals. We discuss their training with particular emphasis on Wiener-space cubature methods based on rough-path analysis as an efficient alternative to Monte Carlo estimation. Together, the three settings address probabilistic prediction, information exchange and emergent behavior in interacting networks, and path-space generative modeling.

---

引用前请核对原文;本文由星际流动整理:https://feed.astralor.com/items/itm_20d485af3f89fdbfc57a8509