Nonparametric LLM Evaluation from Preference Data
This provides practitioners with state-of-the-art methods for comparing or ranking LLMs, addressing a crucial need for LLM leaderboards, but it is incremental as it builds on existing ranking models and debiased machine learning techniques.
The paper tackles the problem of evaluating large language models (LLMs) from human preference data by proposing a nonparametric statistical framework called DMLEval, which uses debiased machine learning to estimate generalized average ranking scores, resulting in statistically efficient estimates and optimal data collection policies.
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.