English

Ethicist: Targeted Training Data Extraction Through Loss Smoothed Soft Prompting and Calibrated Confidence Estimation

Computation and Language 2023-07-11 v1 Artificial Intelligence

Abstract

Large pre-trained language models achieve impressive results across many tasks. However, recent works point out that pre-trained language models may memorize a considerable fraction of their training data, leading to the privacy risk of information leakage. In this paper, we propose a method named Ethicist for targeted training data extraction through loss smoothed soft prompting and calibrated confidence estimation, investigating how to recover the suffix in the training data when given a prefix. To elicit memorization in the attacked model, we tune soft prompt embeddings while keeping the model fixed. We further propose a smoothing loss that smooths the loss distribution of the suffix tokens to make it easier to sample the correct suffix. In order to select the most probable suffix from a collection of sampled suffixes and estimate the prediction confidence, we propose a calibrated confidence estimation method, which normalizes the confidence of the generated suffixes with a local estimation. We show that Ethicist significantly improves the extraction performance on a recently proposed public benchmark. We also investigate several factors influencing the data extraction performance, including decoding strategy, model scale, prefix length, and suffix length. Our code is available at https://github.com/thu-coai/Targeted-Data-Extraction.

Cite

@article{arxiv.2307.04401,
  title  = {Ethicist: Targeted Training Data Extraction Through Loss Smoothed Soft Prompting and Calibrated Confidence Estimation},
  author = {Zhexin Zhang and Jiaxin Wen and Minlie Huang},
  journal= {arXiv preprint arXiv:2307.04401},
  year   = {2023}
}

Comments

ACL 2023 Long Paper (Main Conference)

R2 v1 2026-06-28T11:25:44.643Z