Text Embeddings Reveal (Almost) As Much As Text

John Xavier Morris, Volodymyr Kuleshov, Vitaly Shmatikov, Alexander M Rush

Abstract

How much private information do text embeddings reveal about the original text? We investigate the problem of embedding \textit{inversion}, reconstructing the full text represented in dense text embeddings. We frame the problem as controlled generation: generating text that, when reembedded, is close to a fixed point in latent space. We find that although a naive model conditioned on the embedding performs poorly, a multi-step method that iteratively corrects and re-embeds text is able to recover 92% of 32-token text inputs exactly. We train our model to decode text embeddings from two state-of-the-art embedding models, and also show that our model can recover important personal information (full names) from a dataset of clinical notes.
BibTeX
@inproceedings{morris2023text,
  author = {John Xavier Morris and Volodymyr Kuleshov and Vitaly Shmatikov and Alexander M Rush},
  title = {{Text Embeddings Reveal (Almost) As Much As Text}},
  booktitle = {Conference on Empirical Methods in Natural Language Processing},
  year = {2023},
  doi = {10.18653/v1/2023.emnlp-main.765},
}
Figure