Tracking the perspectives of interacting language models
It addresses the problem of tracking information flow in LLM networks for AI researchers, but is incremental as it builds on existing concepts of model interaction.
The paper formalizes a communication network of large language models (LLMs) and introduces a method to represent individual model perspectives, systematically studying information diffusion in simulated settings.
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.