Saved in:
Bibliographic Details
Main Authors: Xin, Yuan, Li, Zheng, Yu, Ning, Chen, Dingfan, Fritz, Mario, Backes, Michael, Zhang, Yang
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2408.11046
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866929466322714624
author Xin, Yuan
Li, Zheng
Yu, Ning
Chen, Dingfan
Fritz, Mario
Backes, Michael
Zhang, Yang
author_facet Xin, Yuan
Li, Zheng
Yu, Ning
Chen, Dingfan
Fritz, Mario
Backes, Michael
Zhang, Yang
contents Despite being prevalent in the general field of Natural Language Processing (NLP), pre-trained language models inherently carry privacy and copyright concerns due to their nature of training on large-scale web-scraped data. In this paper, we pioneer a systematic exploration of such risks associated with pre-trained language encoders, specifically focusing on the membership leakage of pre-training data exposed through downstream models adapted from pre-trained language encoders-an aspect largely overlooked in existing literature. Our study encompasses comprehensive experiments across four types of pre-trained encoder architectures, three representative downstream tasks, and five benchmark datasets. Intriguingly, our evaluations reveal, for the first time, the existence of membership leakage even when only the black-box output of the downstream model is exposed, highlighting a privacy risk far greater than previously assumed. Alongside, we present in-depth analysis and insights toward guiding future researchers and practitioners in addressing the privacy considerations in developing pre-trained language models.
format Preprint
id arxiv_https___arxiv_org_abs_2408_11046
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Inside the Black Box: Detecting Data Leakage in Pre-trained Language Encoders
Xin, Yuan
Li, Zheng
Yu, Ning
Chen, Dingfan
Fritz, Mario
Backes, Michael
Zhang, Yang
Computation and Language
Despite being prevalent in the general field of Natural Language Processing (NLP), pre-trained language models inherently carry privacy and copyright concerns due to their nature of training on large-scale web-scraped data. In this paper, we pioneer a systematic exploration of such risks associated with pre-trained language encoders, specifically focusing on the membership leakage of pre-training data exposed through downstream models adapted from pre-trained language encoders-an aspect largely overlooked in existing literature. Our study encompasses comprehensive experiments across four types of pre-trained encoder architectures, three representative downstream tasks, and five benchmark datasets. Intriguingly, our evaluations reveal, for the first time, the existence of membership leakage even when only the black-box output of the downstream model is exposed, highlighting a privacy risk far greater than previously assumed. Alongside, we present in-depth analysis and insights toward guiding future researchers and practitioners in addressing the privacy considerations in developing pre-trained language models.
title Inside the Black Box: Detecting Data Leakage in Pre-trained Language Encoders
topic Computation and Language
url https://arxiv.org/abs/2408.11046