# Benchmark Inflation: Revealing LLM Performance Gaps Using Retro-Holdouts

arXiv: https://arxiv.org/abs/2410.09247 · HTML: https://mlsanity.com/paper/2410.09247/

## Key facts

- Cited by 14 papers in the map (data through 14 Oct 2026).
- Builds on 24 of its 45 references that are in the map.
- First posted on arXiv 11 Oct 2024.

## Builds on (most cited first, 24 of 24)

- [BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding](https://mlsanity.com/paper/1810.04805/) (2018)
- [Language Models are Few-Shot Learners](https://mlsanity.com/paper/2005.14165/) (2020)
- [Training Verifiers to Solve Math Word Problems](https://mlsanity.com/paper/2110.14168/) (2021)
- [Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks](https://mlsanity.com/paper/1908.10084/) (2019)
- [Axiomatic Attribution for Deep Networks](https://mlsanity.com/paper/1703.01365/) (2017)
- [TruthfulQA: Measuring How Models Mimic Human Falsehoods](https://mlsanity.com/paper/2109.07958/) (2021)
- [UMAP: Uniform Manifold Approximation and Projection for Dimension Reduction](https://mlsanity.com/paper/1802.03426/) (2018)
- [A Survey on Evaluation of Large Language Models](https://mlsanity.com/paper/2307.03109/) (2023)
- [A Comprehensive Overview of Large Language Models](https://mlsanity.com/paper/2307.06435/) (2023)
- [The WMDP Benchmark: Measuring and Reducing Malicious Use With Unlearning](https://mlsanity.com/paper/2403.03218/) (2024)
- [UltraFeedback: Boosting Language Models with Scaled AI Feedback](https://mlsanity.com/paper/2310.01377/) (2023)
- [Detecting Pretraining Data from Large Language Models](https://mlsanity.com/paper/2310.16789/) (2023)
- [NLP Evaluation in trouble: On the Need to Measure LLM Data Contamination for each Benchmark](https://mlsanity.com/paper/2310.18018/) (2023)
- [Proving Test Set Contamination in Black Box Language Models](https://mlsanity.com/paper/2310.17623/) (2023)
- [A Careful Examination of Large Language Model Performance on Grade School Arithmetic](https://mlsanity.com/paper/2405.00332/) (2024)
- [Time Travel in LLMs: Tracing Data Contamination in Large Language Models](https://mlsanity.com/paper/2308.08493/) (2023)
- [Investigating Data Contamination in Modern Benchmarks for Large Language Models](https://mlsanity.com/paper/2311.09783/) (2023)
- [When Benchmarks are Targets: Revealing the Sensitivity of Large Language Model Leaderboards](https://mlsanity.com/paper/2402.01781/) (2024)
- [Investigating Data Contamination for Pre-training Language Models](https://mlsanity.com/paper/2401.06059/) (2024)
- [The Foundation Model Transparency Index](https://mlsanity.com/paper/2310.12941/) (2023)
- [LatestEval: Addressing Data Contamination in Language Model Evaluation through Dynamic and Time-Sensitive Test Construction](https://mlsanity.com/paper/2312.12343/) (2023)
- [Goodhart's Law in Reinforcement Learning](https://mlsanity.com/paper/2310.09144/) (2023)
- [Pretraining on the Test Set Is All You Need](https://mlsanity.com/paper/2309.08632/) (2023)
- [On the term "randomization test"](https://mlsanity.com/paper/2306.07720/) (2023)

## Built on by (most cited first, 14 of 14)

- [AIRS-Bench: a Suite of Tasks for Frontier AI Research Science Agents](https://mlsanity.com/paper/2602.06855/) (2026)
- [Reward Hacking Benchmark: Measuring Exploits in LLM Agents with Tool Use](https://mlsanity.com/paper/2605.02964/) (2026)
- [DarkBench: Benchmarking Dark Patterns in Large Language Models](https://mlsanity.com/paper/2503.10728/) (2025)
- [Does Data Contamination Detection Work (Well) for LLMs? A Survey and Evaluation on Detection Assumptions](https://mlsanity.com/paper/2410.18966/) (2024)
- [SEA-HELM: Southeast Asian Holistic Evaluation of Language Models](https://mlsanity.com/paper/2502.14301/) (2025)
- [The Emperor's New Clothes in Benchmarking? A Rigorous Examination of Mitigation Strategies for LLM Benchmark Data Contamination](https://mlsanity.com/paper/2503.16402/) (2025)
- [Risk Management for Mitigating Benchmark Failure Modes: BenchRisk](https://mlsanity.com/paper/2510.21460/) (2025)
- [Engineering Reliable Coding Agents: Evaluating and Operating the System Around the Model](https://mlsanity.com/paper/2608.13867/) (2026)
- [Temporal Leakage in LLM Backtesting: Measurement, Validation, and Adjusted Scores](https://mlsanity.com/paper/2608.02985/) (2026)
- [Adversarial Pragmatics for AI Safety Evaluation: A Diagnostic Framework and Seed Benchmark for Language-Mediated Control](https://mlsanity.com/paper/2607.01153/) (2026)
- [Beyond Success Rate: Cost-Aware Evaluation of Offensive and Defensive Security Agents](https://mlsanity.com/paper/2607.15263/) (2026)
- [Selecting Language Models for Social Science: Start Small, Start Open, and Validate](https://mlsanity.com/paper/2601.10926/) (2026)
- [Benchmark Health Index: A Systematic Framework for Benchmarking the Benchmarks of LLMs](https://mlsanity.com/paper/2602.11674/) (2026)
- [The AI Evaluation Ecosystem](https://mlsanity.com/paper/2610.09296/) (2026)

## Abstract (arXiv)

The training data for many Large Language Models (LLMs) is contaminated with test data. This means that public benchmarks used to assess LLMs are compromised, suggesting a performance gap between benchmark scores and actual capabilities. Ideally, a private holdout set could be used to accurately verify scores. Unfortunately, such datasets do not exist for most benchmarks, and post-hoc construction of sufficiently similar datasets is non-trivial. To address these issues, we introduce a systematic methodology for (i) retrospectively constructing a holdout dataset for a target dataset, (ii) demonstrating the statistical indistinguishability of this retro-holdout dataset, and (iii) comparing LLMs on the two datasets to quantify the performance gap due to the dataset's public availability. Applying these methods to TruthfulQA, we construct and release Retro-Misconceptions, on which we evaluate twenty LLMs and find that some have inflated scores by as much as 16 percentage points. Our results demonstrate that public benchmark scores do not always accurately assess model properties, and underscore the importance of improved data practices in the field.