Beyond Memorization: Violating Privacy Via Inference with Large Language Models

arXiv:2310.07298v2 Announce Type: replace-cross
Abstract: Current privacy research on large language models (LLMs) primarily focuses on the issue of extracting memorized training data. At the same time, models' inference capabilities have increased drastically. This raises the key question of whether current LLMs could violate individuals' privacy by inferring personal attributes from text given at inference time. In this work, we present the first comprehensive study on the capabilities of pretrained LLMs to infer personal attributes from text. We construct a …

abstract arxiv beyond capabilities cs.ai cs.lg current data inference issue key language language models large language large language models llms privacy question raises research the key training training data type via