LLM Comparator: Visual Analytics for Side-by-Side Evaluation of Large Language Models

TL;DR

LLM Comparator is a visual analytics tool designed to help researchers and engineers interpret and analyze side-by-side evaluations of large language models, addressing scalability and interpretability challenges.

Contribution

The paper introduces LLM Comparator, a novel interactive visualization tool specifically built for analyzing and understanding LLM evaluation results.

Findings

Enhanced understanding of model performance differences

Improved interpretability of evaluation results

Validated effectiveness through user study

Abstract

Automatic side-by-side evaluation has emerged as a promising approach to evaluating the quality of responses from large language models (LLMs). However, analyzing the results from this evaluation approach raises scalability and interpretability challenges. In this paper, we present LLM Comparator, a novel visual analytics tool for interactively analyzing results from automatic side-by-side evaluation. The tool supports interactive workflows for users to understand when and why a model performs better or worse than a baseline model, and how the responses from two models are qualitatively different. We iteratively designed and developed the tool by closely working with researchers and engineers at a large technology company. This paper details the user challenges we identified, the design and development of the tool, and an observational study with participants who regularly evaluate their models.