Examining the robustness of LLM evaluation to the distributional assumptions of benchmarks

2024/04/25 by Ailem, Melissa, Marazopoulou, Katerina, Siska, Charlotte +1 · 5 citations
#Computation and Language (cs.CL) #FOS: Computer and information sciences

paper · doi:10.48550/arxiv.2404.16966

Abstract

Benchmarks have emerged as the central approach for evaluating Large Language Models (LLMs). The research community often relies on a model's average performance across the test prompts of a benchmark to evaluate the model's performance. This is consistent with the assumption that the test prompts within a benchmark represent a random sample from a real-world distribution of interest. We note that this is generally not the case; instead, we hold that the distribution of interest varies according to the specific use case. We find that (1) the correlation in model performance across test prompts is non-random, (2) accounting for correlations across test prompts can change model rankings on major benchmarks, (3) explanatory factors for these correlations include semantic similarity and common LLM failure points.

Cited by

• A Principle-based Framework for the Development and Evaluation of Large Language Models for Health and Wellness
• Structure-aware Propagation Generation with Large Language Models for Fake News Detection
• Correcting Prompt Dependence in LLM Benchmarks: A Bayesian Hierarchical Model with Embedding-Space Clustering
• Performance Assessment Strategies for Generative AI Applications in Healthcare
• Guidelines for Empirical Studies in Software Engineering involving Large Language Models

Related

• On Robustness and Reliability of Benchmark-Based Evaluation of LLMs
• Towards Reproducible LLM Evaluation: Quantifying Uncertainty in LLM Benchmark Scores
• Benchmark Inflation: Revealing LLM Performance Gaps Using Retro-Holdouts
• The Vulnerability of Language Model Benchmarks: Do They Accurately Reflect True LLM Performance?
• How Benchmark Prediction from Fewer Data Misses the Mark