# Detecting Data Contamination in Large Language Models

arXiv: https://arxiv.org/abs/2604.19561 · HTML: https://mlsanity.com/paper/2604.19561/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 11 of its 19 references that are in the map.
- First posted on arXiv 21 Apr 2026.

## Builds on (most cited first, 11 of 11)

- [Extracting Training Data from Diffusion Models](https://mlsanity.com/paper/2301.13188/) (2023)
- [Proving Test Set Contamination in Black Box Language Models](https://mlsanity.com/paper/2310.17623/) (2023)
- [On Memorization of Large Language Models in Logical Reasoning](https://mlsanity.com/paper/2410.23123/) (2024)
- [DE-COP: Detecting Copyrighted Content in Language Models Training Data](https://mlsanity.com/paper/2402.09910/) (2024)
- [ReCaLL: Membership Inference via Relative Conditional Log-Likelihoods](https://mlsanity.com/paper/2406.15968/) (2024)
- [Localizing Paragraph Memorization in Language Models](https://mlsanity.com/paper/2403.19851/) (2024)
- [Undesirable Memorization in Large Language Models: A Survey](https://mlsanity.com/paper/2410.02650/) (2024)
- [Does Data Contamination Detection Work (Well) for LLMs? A Survey and Evaluation on Detection Assumptions](https://mlsanity.com/paper/2410.18966/) (2024)
- [Con-ReCall: Detecting Pre-training Data in LLMs via Contrastive Decoding](https://mlsanity.com/paper/2409.03363/) (2024)
- [Detecting Training Data of Large Language Models via Expectation Maximization](https://mlsanity.com/paper/2410.07582/) (2024)
- [Adaptive Pre-training Data Detection for Large Language Models via Surprising Tokens](https://mlsanity.com/paper/2407.21248/) (2024)

## Abstract (arXiv)

Large Language Models (LLMs) utilize large amounts of data for their training, some of which may come from copyrighted sources. Membership Inference Attacks (MIA) aim to detect those documents and whether they have been included in the training corpora of the LLMs. The black-box MIAs require a significant amount of data manipulation; therefore, their comparison is often challenging. We study state-of-the-art (SOTA) MIAs under the black-box assumptions and compare them to each other using a unified set of datasets to determine if any of them can reliably detect membership under SOTA LLMs. In addition, a new method, called the Familiarity Ranking, was developed to showcase a possible approach to black-box MIAs, thereby giving LLMs more freedom in their expression to understand their reasoning better. The results indicate that none of the methods are capable of reliably detecting membership in LLMs, as shown by an AUC-ROC of approximately 0.5 for all methods across several LLMs. The higher TPR and FPR for more advanced LLMs indicate higher reasoning and generalizing capabilities, showcasing the difficulty of detecting membership in LLMs using black-box MIAs.