SoK: Membership Inference Attacks on LLMs are Rushing Nowhere (and How to Fix It)

• URL: http://arxiv.org/abs/2406.17975v2
• Date: Mon, 07 Oct 2024 17:49:13 GMT
• Title: SoK: Membership Inference Attacks on LLMs are Rushing Nowhere (and How to Fix It)
• Authors: Matthieu Meeus, Igor Shilov, Shubham Jain, Manuel Faysse, Marek Rei, Yves-Alexandre de Montjoye,
• Abstract summary: More than 10 new methods have been proposed to perform Membership Inference Attacks (MIAs) against LLMs. Contrary to traditional MIAs which rely on fixed -- but randomized -- records or models, these methods are mostly evaluated on datasets collected post-hoc. This lack of randomization raises concerns of a distribution shift between members and non-members.
• Score: 16.673210422615348
• License:
• Abstract: Whether LLMs memorize their training data and what this means, from privacy leakage to detecting copyright violations -- has become a rapidly growing area of research over the last two years. In recent months, more than 10 new methods have been proposed to perform Membership Inference Attacks (MIAs) against LLMs. Contrary to traditional MIAs which rely on fixed -- but randomized -- records or models, these methods are mostly evaluated on datasets collected post-hoc. Sets of members and non-members, used to evaluate the MIA, are constructed using informed guesses after the release of a model. This lack of randomization raises concerns of a distribution shift between members and non-members. In the first part, we review the literature on MIAs against LLMs. While most work focuses on sequence-level MIAs evaluated in post-hoc setups, we show that a range of target models, motivations and units of interest have been considered in the literature. We then quantify distribution shifts present in the 6 datasets used in the literature, ranging from books to papers, using a bag of word classifier. Our analysis reveals that all of them suffer from severe distribution shifts. This challenges the validity of using such setups to measure LLM memorization and may undermine the benchmarking of recently proposed methods. Yet, all hope might not be lost. In the second part, we introduce important considerations to properly evaluate MIAs against LLMs and discuss potential ways forward: randomized test splits, injections of randomized (unique) sequences, randomized finetuning, and post-hoc control methods. While each option comes with its advantages and limitations, we believe they collectively provide solid grounds to guide the development of MIA methods and study LLM memorization. We conclude by proposing comprehensive, easy-to-use benchmarks for sequence- and document-level MIAs against LLMs.

Related papers

• Scaling Up Membership Inference: When and How Attacks Succeed on Large Language Models [37.420266437306374]
  Membership inference attacks (MIA) attempt to verify the membership of a given data sample in the training set for a model. Recent research has largely concluded that current MIA methods do not work on large language models (LLM)
  arXiv  Detail & Related papers  (2024-10-31T18:59:46Z)
• Beyond Binary: Towards Fine-Grained LLM-Generated Text Detection via Role Recognition and Involvement Measurement [51.601916604301685]
  Large language models (LLMs) generate content that can undermine trust in online discourse. Current methods often focus on binary classification, failing to address the complexities of real-world scenarios like human-AI collaboration. To move beyond binary classification and address these challenges, we propose a new paradigm for detecting LLM-generated content.
  arXiv  Detail & Related papers  (2024-10-18T08:14:10Z)
• Detecting Training Data of Large Language Models via Expectation Maximization [62.28028046993391]
  Membership inference attacks (MIAs) aim to determine whether a specific instance was part of a target model's training data. Applying MIAs to large language models (LLMs) presents unique challenges due to the massive scale of pre-training data and the ambiguous nature of membership. We introduce EM-MIA, a novel MIA method for LLMs that iteratively refines membership scores and prefix scores via an expectation-maximization algorithm.
  arXiv  Detail & Related papers  (2024-10-10T03:31:16Z)
• MIA-Tuner: Adapting Large Language Models as Pre-training Text Detector [32.15773300068426]
  Existing methods have designed various sophisticated MIA score functions to achieve considerable detection performance. We propose MIA-Tuner, a novel instruction-based MIA method, which instructs LLMs themselves to serve as a more precise pre-training data detector. We design two instruction-based safeguards to respectively mitigate the privacy risks brought by the existing methods and MIA-Tuner.
  arXiv  Detail & Related papers  (2024-08-16T11:09:56Z)
• Cycles of Thought: Measuring LLM Confidence through Stable Explanations [53.15438489398938]
  Large language models (LLMs) can reach and even surpass human-level accuracy on a variety of benchmarks, but their overconfidence in incorrect responses is still a well-documented failure mode. We propose a framework for measuring an LLM's uncertainty with respect to the distribution of generated explanations for an answer.
  arXiv  Detail & Related papers  (2024-06-05T16:35:30Z)
• Decompose and Aggregate: A Step-by-Step Interpretable Evaluation Framework [75.81096662788254]
  Large Language Models (LLMs) are scalable and economical evaluators. The question of how reliable these evaluators are has emerged as a crucial research question. We propose Decompose and Aggregate, which breaks down the evaluation process into different stages based on pedagogical practices.
  arXiv  Detail & Related papers  (2024-05-24T08:12:30Z)
• Do Membership Inference Attacks Work on Large Language Models? [141.2019867466968]
  Membership inference attacks (MIAs) attempt to predict whether a particular datapoint is a member of a target model's training data. We perform a large-scale evaluation of MIAs over a suite of language models trained on the Pile, ranging from 160M to 12B parameters. We find that MIAs barely outperform random guessing for most settings across varying LLM sizes and domains.
  arXiv  Detail & Related papers  (2024-02-12T17:52:05Z)
• Practical Membership Inference Attacks against Fine-tuned Large Language Models via Self-prompt Calibration [32.15773300068426]
  Membership Inference Attacks (MIAs) aim to infer whether a target data record has been utilized for model training or not. We propose a Membership Inference Attack based on Self-calibrated Probabilistic Variation (SPV-MIA) Specifically, since memorization in LLMs is inevitable during the training process and occurs before overfitting, we introduce a more reliable membership signal.
  arXiv  Detail & Related papers  (2023-11-10T13:55:05Z)
• Fake Alignment: Are LLMs Really Aligned Well? [91.26543768665778]
  This study investigates the substantial discrepancy in performance between multiple-choice questions and open-ended questions. Inspired by research on jailbreak attack patterns, we argue this is caused by mismatched generalization.
  arXiv  Detail & Related papers  (2023-11-10T08:01:23Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.