mlsanityData to 7 Oct 2026

Semantic Membership Inference Attack against Large Language Models

arXiv 2406.10218 ↗First posted 14 Jun 2024cs.LGcited by 17

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

Abstract (arXiv)

Membership Inference Attacks (MIAs) determine whether a specific data point was included in the training set of a target model. In this paper, we introduce the Semantic Membership Inference Attack (SMIA), a novel approach that enhances MIA performance by leveraging the semantic content of inputs and their perturbations. SMIA trains a neural network to analyze the target model's behavior on perturbed inputs, effectively capturing variations in output probability distributions between members and non-members. We conduct comprehensive evaluations on the Pythia and GPT-Neo model families using the Wikipedia dataset. Our results show that SMIA significantly outperforms existing MIAs; for instance, SMIA achieves an AUC-ROC of 67.39% on Pythia-12B, compared to 58.90% by the second-best attack.

Key facts. Cited by 17 papers in the map (data through 7 Oct 2026). Placed in Membership inference attacks, one of 12 directions in LLM safety and security. Builds on 26 of its 35 references that are in the map. First posted on arXiv 14 Jun 2024.

Builds on

26 of 35 references.

Closest ancestors · 4

Same direction, newest first

  1. 2024Do Membership Inference Attacks Work on Large Language Models?12 Feb
  2. 2023Membership Inference Attacks against Language Models via Neighbourhood Comparison29 May
  3. 2022Quantifying Privacy Risks of Masked Language Models Using Membership Inference Attacks8 Mar
  4. 2021Membership Inference Attack Susceptibility of Clinical Language Models16 Apr

Other references · 22

  1. 2018BERT: Pre-training of Deep Bidirectional Transformers for Language Understandingcited by 40,204
  2. 2019Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformercited by 14,491
  3. 2022PaLM: Scaling Language Modeling with Pathwayscited by 5,304
  4. 2021Finetuned Language Models Are Zero-Shot Learnerscited by 3,747
  5. 2020The Pile: An 800GB Dataset of Diverse Text for Language Modelingcited by 2,346
  6. 2016Membership Inference Attacks against Machine Learning Modelscited by 2,221
  7. 2020Extracting Training Data from Large Language Modelscited by 2,144
  8. 2022Survey of Hallucination in Natural Language Generationcited by 2,056
  9. 2023Pythia: A Suite for Analyzing Large Language Models Across Training and Scalingcited by 1,734
  10. 2022Quantifying Memorization Across Neural Language Modelscited by 755
  11. 2023DetectGPT: Zero-Shot Machine-Generated Text Detection using Probability Curvaturecited by 705
  12. 2023Scalable Extraction of Training Data from (Production) Language Modelscited by 415
10 more
  1. 2023Detecting Pretraining Data from Large Language Modelscited by 375
  2. 2023Who's Harry Potter? Approximate Unlearning in LLMscited by 366
  3. 2024Do Large Language Models Latently Perform Multi-Hop Reasoning?cited by 165
  4. 2021Counterfactual Memorization in Neural Language Modelscited by 155
  5. 2022The Privacy Onion Effect: Memorization is Relativecited by 128
  6. 2023Do Language Models Know When They're Hallucinating References?cited by 115
  7. 2024Min-K%++: Improved Baseline for Detecting Pre-Training Data from Large Language Modelscited by 102
  8. 2021Does BERT Pretrained on Clinical Notes Reveal Sensitive Data?cited by 91
  9. 2024Why Does Differential Privacy with Large Epsilon Defend Against Practical Membership Inference Attacks?cited by 12
  10. 2019Assessing differentially private deep learning with Membership Inferencecited by 11

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2024: 2; 2025: 9; 2026: 6.

17 papers.

Sort:

By area:

  1. 2025A Comprehensive Survey in LLM(-Agent) Full Stack Safety: Data, Training and Deploymentcited by 110
  2. 2024SoK: Membership Inference Attacks on LLMs are Rushing Nowhere (and How to Fix It)cited by 54
  3. 2025Membership Inference Attacks on Large-Scale Models: A Surveycited by 17
  4. 2024Detecting Training Data of Large Language Models via Expectation Maximizationcited by 13
  5. 2025Permissioned LLMs: Enforcing Access Control in Large Language Modelscited by 7
  6. 2026Window-based Membership Inference Attacks Against Fine-tuned Large Language Modelscited by 5
  7. 2026Membership Inference Attacks Against Fine-tuned Diffusion Language Modelscited by 5
  8. 2026AttenMIA: LLM Membership Inference Attack through Attention Signalscited by 4
  9. 2025Quantifying the Privacy Implications of High-Fidelity Synthetic Network Trafficcited by 3
  10. 2025Memories Retrieved from Many Paths: A Multi-Prefix Framework for Robust Detection of Training Data Leakage in Large Language Modelscited by 2
  11. 2025LLM in the Middle: A Systematic Review of Threats and Mitigations to Real-World LLM-based Systemscited by 1
  12. 2025Multimodal Evaluation of Russian-language Architecturescited by 1
5 more
  1. 2026AI Security in the Foundation Model Era: A Comprehensive Survey from a Unified Perspectivecited by 1
  2. 2026Pretraining Data Exposure in Large Language Models: A Survey of Membership Inference, Data Contamination, and Security Implicationscited by 1
  3. 2025EL-MIA: Quantifying Membership Inference Risks of Sensitive Entities in LLMscited by 0
  4. 2025FiMMIA: scaling semantic perturbation-based membership inference across modalitiescited by 0
  5. 2026Calibrating One-Round Membership Inference with Neighborscited by 0