Tracking the perspectives of interacting language models

2024/06/17 by Hayden S. Helm, Helm, Hayden, Brandon Duderstadt +5 · 2 citations
Computer Science · #Natural Language Processing Techniques

paper · pdf · doi:10.48550/arxiv.2406.11938

Abstract

Large language models (LLMs) are capable of producing high quality information at unprecedented rates. As these models continue to entrench themselves in society, the content they produce will become increasingly pervasive in databases that are, in turn, incorporated into the pre-training data, fine-tuning data, retrieval data, etc. of other language models. In this paper we formalize the idea of a communication network of LLMs and introduce a method for representing the perspective of individual models within a collection of LLMs. Given these tools we systematically study information diffusion in the communication network of LLMs in various simulated settings.

Cited by

• Detecting Perspective Shifts in Multi-agent Systems
• Concentration bounds on response-based vector embeddings of black-box generative models

Related

• Information Diffusion and Preferential Attachment in a Network of Large Language Models
• Complex networks and human language
• Learnings from Data Integration for Augmented Language Models
• Network Formation and Dynamics Among Multi-LLMs
• LLM Web Dynamics: Tracing Model Collapse in a Network of LLMs