BenchMarker: An Education-Inspired Toolkit for Highlighting Flaws in Multiple-Choice Benchmarks

TL;DR

BenchMarker is a toolkit inspired by education research that uses LLM judges to identify common flaws in multiple-choice benchmarks, revealing persistent issues that affect NLP evaluation accuracy.

Contribution

Introduces BenchMarker, a novel education-inspired toolkit leveraging LLMs to detect contamination, shortcuts, and writing errors in MCQ benchmarks, improving quality control.

Findings

47% of TruthfulQA questions appear online, indicating contamination.

100% of HellaSwag questions violate at least one writing rule.

Flaws in MCQs can inflate or deflate model accuracy and alter benchmark rankings.

Abstract

Multiple-choice question answering (MCQA) is standard in NLP, but benchmarks lack rigorous quality control. We present BenchMarker, an education-inspired toolkit using LLM judges to flag three common MCQ flaws: 1) contamination: items appearing exactly online; 2) shortcuts: cues in the choices that enable guessing; and 3) writing errors: structural/grammatical issues based on a 19-rule education rubric. We validate BenchMarker with human annotations, then run the tool to audit 12 benchmarks, revealing: 1) flaws persist in MCQA benchmarks, especially automatically-made and crowdsourced data - we detect 47% of TruthfulQA appears online and 100% of HellaSwag violates multiple writing rules; 2) contaminated MCQs tend to inflate accuracy, while writing errors tend to lower it and change rankings beyond random; and 3) prior benchmark repairs address their targeted issues (i.e., lowering accuracy with LLM-written distractors), but inadvertently add new flaws (i.e. implausible distractors, many correct answers). Overall, flaws in MCQs degrade NLP evaluation, but education research offers a path forward. We release BenchMarker to bridge the fields and improve MCQA benchmark design.