Nonparametric LLM Evaluation from Preference Data

TL;DR

This paper introduces DMLEval, a nonparametric framework for evaluating and ranking large language models using preference data, addressing limitations of existing methods by providing efficient, flexible, and theoretically sound tools.

Contribution

The paper presents DMLEval, a novel nonparametric approach that leverages debiased machine learning to improve LLM evaluation and ranking from human preferences.

Findings

DMLEval produces statistically efficient LLM rankings.

It effectively incorporates complex human response data.

The framework is validated on synthetic and real datasets.

Abstract

Evaluating the performance of large language models (LLMs) from human preference data is crucial for obtaining LLM leaderboards. However, many existing approaches either rely on restrictive parametric assumptions or lack valid uncertainty quantification when flexible machine learning methods are used. In this paper, we propose a nonparametric statistical framework, DMLEval, for comparing and ranking LLMs from preference data using debiased machine learning (DML). For this, we introduce generalized average ranking scores (GARS), which generalize commonly used ranking models, including the Bradley-Terry model or PageRank/ Rank centrality, with complex human responses such as ties. DMLEval comes with the following advantages: (i) It produces statistically efficient estimates of GARS ranking scores. (ii) It naturally allows the incorporation of black-box machine learning methods for estimation. (iii) It can be combined with pre-trained LLM evaluators (e.g., using LLM-as-a-judge). (iv) It suggests optimal policies for collecting preference data under budget constraints. We demonstrate these advantages both theoretically and empirically using both synthetic and real-world preference datasets. In summary, our framework provides practitioners with powerful, state-of-the-art methods for comparing or ranking LLMs.