mlsanityData to 14 Oct 2026

Tab-MIA: A Benchmark Dataset for Membership Inference Attacks on Tabular Data in LLMs

arXiv 2507.17259 ↗First posted 23 Jul 2025cs.CR cs.CLcited by 3

On the map: LLM safety and security → Membership inference attacks

Abstract (arXiv)

Large language models (LLMs) are increasingly trained on tabular data, which, unlike unstructured text, often contains personally identifiable information (PII) in a highly structured and explicit format. As a result, privacy risks arise, since sensitive records can be inadvertently retained by the model and exposed through data extraction or membership inference attacks (MIAs). While existing MIA methods primarily target textual content, their efficacy and threat implications may differ when applied to structured data, due to its limited content, diverse data types, unique value distributions, and column-level semantics. In this paper, we present Tab-MIA, a benchmark dataset for evaluating MIAs on tabular data in LLMs and demonstrate how it can be used. Tab-MIA comprises five data collections, each represented in six different encoding formats. Using our Tab-MIA benchmark, we conduct the first evaluation of state-of-the-art MIA methods on LLMs finetuned with tabular data across multiple encoding formats. In the evaluation, we analyze the memorization behavior of pretrained LLMs on structured data derived from Wikipedia tables. Our findings show that LLMs memorize tabular data in ways that vary across encoding formats, making them susceptible to extraction via MIAs. Even when fine-tuned for as few as three epochs, models exhibit high vulnerability, with AUROC scores approaching 90% in most cases. Tab-MIA enables systematic evaluation of these risks and provides a foundation for developing privacy-preserving methods for tabular data in LLMs.

Key facts. Cited by 3 papers in the map (data through 14 Oct 2026). Placed in Membership inference attacks, one of 12 directions in LLM safety and security. Builds on 32 of its 47 references that are in the map. First posted on arXiv 23 Jul 2025.

Builds on

32 of 47 references.

Closest ancestors · 5

Same direction, newest first

  1. 2025Tag&Tab: Pretraining Data Detection in Large Language Models Using Keyword-Based Membership Inference Attack14 Jan
  2. 2024EM-MIAs: Enhancing Membership Inference Attacks in Large Language Models through Ensemble Modeling23 Dec
  3. 2023Analyzing Leakage of Personally Identifiable Information in Language Models1 Feb
  4. 2021Membership Inference Attacks From First Principles7 Dec
  5. 2021Membership Inference Attacks on Machine Learning: A Survey14 Mar

Other references · 27

  1. 2016XGBoost: A Scalable Tree Boosting Systemcited by 5,166
  2. 2023QLoRA: Efficient Finetuning of Quantized LLMscited by 3,418
  3. 2016Membership Inference Attacks against Machine Learning Modelscited by 2,281
  4. 2020Extracting Training Data from Large Language Modelscited by 2,189
  5. 2022Quantifying Memorization Across Neural Language Modelscited by 763
  6. 2021Revisiting Deep Learning Models for Tabular Datacited by 740
  7. 2015Compositional Semantic Parsing on Semi-Structured Tablescited by 515
  8. 2021Deep Neural Networks and Tabular Data: A Surveycited by 491
  9. 2020TAPAS: Weakly Supervised Table Parsing via Pre-trainingcited by 434
  10. 2019TabFact: A Large-scale Dataset for Table-based Fact Verificationcited by 428
  11. 2020TaBERT: Pretraining for Joint Understanding of Textual and Tabular Datacited by 398
  12. 2023Detecting Pretraining Data from Large Language Modelscited by 387
15 more
  1. 2022Language Models are Realistic Tabular Data Generatorscited by 294
  2. 2022TabLLM: Few-shot Classification of Tabular Data with Large Language Modelscited by 277
  3. 2023A Comprehensive Survey of AI-Generated Content (AIGC): A History of Generative AI from GAN to ChatGPTcited by 224
  4. 2024Large Language Models(LLMs) on Tabular Data: Prediction, Generation, and Understanding -- A Surveycited by 159
  5. 2022LIFT: Language-Interfaced Fine-Tuning for Non-Language Machine Learning Taskscited by 148
  6. 2022Large Language Models are few(1)-shot Table Reasonerscited by 116
  7. 2023Table Meets LLM: Can Large Language Models Understand Structured Table Data? A Benchmark and Empirical Studycited by 116
  8. 2023Tabular Representation, Noisy Operators, and Impacts on Table Structure Understanding Tasks in LLMscited by 49
  9. 2023Exploring Memorization in Fine-tuned Language Modelscited by 45
  10. 2023TAP4LLM: Table Provider on Sampling, Augmenting, and Packing Semi-structured Data for Large Language Model Reasoningcited by 32
  11. 2023TABLET: Learning From Instructions For Tabular Datacited by 28
  12. 2024NormTab: Improving Symbolic Reasoning in LLMs Through Tabular Data Normalizationcited by 16
  13. 2025Optimizing Pretraining Data Mixtures with LLM-Estimated Utilitycited by 15
  14. 2023Towards Better Serialization of Tabular Data for Few-shot Classification with Large Language Modelscited by 12
  15. 2023JarviX: A LLM No code Platform for Tabular Data Analysis and Optimizationcited by 10

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 3.

3 papers.

  1. 2026Toward Reliable, Safe, and Secure LLMs for Scientific Applicationscited by 1
  2. 2026Privacy Vulnerabilities of Attention Layers in Tabular Foundation Models and Protection of High-Risk Queriescited by 1
  3. 2026Probing Memorization of Tabular In-Context Learningcited by 1

Similar, not cited

Close in wording, no citation either way

  1. 2025When Tables Leak: Attacking String Memorization in LLM-Based Tabular Data Generation
  2. 2025On the Effectiveness of Membership Inference in Targeted Data Extraction from Large Language Models
  3. 2024Order of Magnitude Speedups for LLM Membership Inference
  4. 2024MIA-Tuner: Adapting Large Language Models as Pre-training Text Detector
  5. 2026Membership Inference on LLMs in the Wild
  6. 2025Membership Inference Attacks on Large-Scale Models: A Survey
  7. 2024SoK: Membership Inference Attacks on LLMs are Rushing Nowhere (and How to Fix It)
  8. 2026What Hard Tokens Reveal: Exploiting Low-confidence Tokens for Membership Inference Attacks against Large Language Models