Training Data Extraction From Pre-trained Language Models: A Survey

Shotaro Ishihara
5/25/2023
cs.CL

No confident code implementation found for this paper. 1 possible match can be shown below.

Abstract

As the deployment of pre-trained language models (PLMs) expands, pressing security concerns have arisen regarding the potential for malicious extraction of training data, posing a threat to data privacy. This study is the first to provide a comprehensive survey of training data extraction from PLMs. Our review covers more than 100 key papers in fields such as natural language processing and security. First, preliminary knowledge is recapped and a taxonomy of various definitions of memorization is presented. The approaches for attack and defense are then systemized. Furthermore, the empirical findings of several quantitative studies are highlighted. Finally, future research directions based on this review are suggested.

View on arXivView PDF

Code Implementations

No confident code match yet

We couldn't find an author-owned or strongly-evidenced community implementation for this paper. 1 weaker match is hidden by default — verify before relying on them.

No code implementations found yet.

Know of an implementation? Let us know in the comments below!

Similar Papers

New Emerged Security and Privacy of Pre-trained Model: a Survey and Outlook

89% match1 repo12024

Traces of Memorisation in Large Language Models for Code

88% match1 repo52023

Persistent Pre-Training Poisoning of LLMs

87% match1 repo132024

Extracting Training Data from Large Language Models

86% match1 repo1982020

A Survey on Data Security in Large Language Models

86% match2025

Identifying and Mitigating Privacy Risks Stemming from Language Models: A Survey

86% match2023

Cite this paper

@article{ishihara2023training,
  title  = {Training Data Extraction From Pre-trained Language Models: A Survey},
  author = {Shotaro Ishihara},
  year   = {2023},
  eprint = {2305.16157},
  archivePrefix = {arXiv},
  url    = {http://arxiv.org/abs/2305.16157v1}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/2305.16157.svg)](https://codeofpaper.com/paper/2305.16157)

Discussion