# MMLU-Pro Contamination Audit

> MMLU-Pro scores 1.1/10 (Tier D, Compromised) on BenchAudit as of November 2026, reflecting publicly downloadable plaintext test files and standardized independent third-party grading.

- **Canonical URL:** https://benchaudit.org/benchmarks/mmlu-pro/
- **Benchmark ID:** `mmlu-pro`
- **Hub:** TIGER-AI-Lab / Artificial Analysis
- **Category:** General
- **Domains:** Science, Math, Business, Legal, Medical, Humanities
- **Skills:** Reasoning, Knowledge Work
- **BenchAudit Score:** 1.1/10 (Tier D - Compromised)
- **Last Audited:** 2026-11-17 (Ref 2026-11)

## Factor Scores

| Factor | Score (/10) | Weight |
| --- | ---: | ---: |
| Privacy | 0 | 40% |
| Originality | 0 | 25% |
| Freshness | 0 | 20% |
| Grading | 7.3 | 15% |

## Primary-Source Evidence

### Privacy
> "MMLU-Pro dataset is a more robust and challenging massive multi-task understanding dataset tailored to more rigorously benchmark large language models' capabilities. This dataset contains 12K complex questions across various disciplines."
- Source: https://huggingface.co/datasets/TIGER-Lab/MMLU-Pro

### Originality
> "MMLU-Pro comprises over 12,000 rigorously curated questions from academic exams and textbooks, spanning 14 diverse domains including Biology, Business, Chemistry, Computer Science, Economics, Engineering, Health, History, Law, Math, Philosophy, Physics, Psychology, and Others."
- Source: https://github.com/TIGER-AI-Lab/MMLU-Pro

### Freshness
> "arXiv:2406.01574 (cs) [Submitted on 3 Jun 2024 ( v1 ), last revised 6 Nov 2024 (this version, v6)]"
- Source: https://arxiv.org/abs/2406.01574

### Grading
> "10 option multiple choice format Regex-based answer extraction with pass@1 scoring (prompt and regex below)"
- Source: https://artificialanalysis.ai/methodology/intelligence-benchmarking
