PhyloLM : Inferring the Phylogeny of Large Language Models and Predicting their Performances in Benchmarks

TL;DR

This paper presents PhyloLM, a novel method that uses phylogenetic algorithms to analyze relationships among large language models and predict their benchmark performances, aiding in understanding LLM evolution without needing training details.

Contribution

Introduces PhyloLM, adapting phylogenetic techniques to LLMs for relationship mapping and performance prediction, a novel approach in LLM analysis.

Findings

PhyloLM accurately captures known relationships among LLMs.

The phylogenetic distance predicts benchmark performance effectively.

The method works with open-source and closed models.

Abstract

This paper introduces PhyloLM, a method adapting phylogenetic algorithms to Large Language Models (LLMs) to explore whether and how they relate to each other and to predict their performance characteristics. Our method calculates a phylogenetic distance metric based on the similarity of LLMs' output. The resulting metric is then used to construct dendrograms, which satisfactorily capture known relationships across a set of 111 open-source and 45 closed models. Furthermore, our phylogenetic distance predicts performance in standard benchmarks, thus demonstrating its functional validity and paving the way for a time and cost-effective estimation of LLM capabilities. To sum up, by translating population genetic concepts to machine learning, we propose and validate a tool to evaluate LLM development, relationships and capabilities, even in the absence of transparent training information.