Skip to Main content Skip to Navigation
Conference papers

Investigating the Impact of Pre-trained Word Embeddings on Memorization in Neural Networks

Abstract : The sensitive information present in the training data, poses a privacy concern for applications as their unintended memorization during training can make models susceptible to membership inference and attribute inference attacks. In this paper, we investigate this problem in various pre-trained word embeddings (GloVe, ELMo and BERT) with the help of language models built on top of it. In particular, firstly sequences containing sensitive information like a single-word disease and 4-digit PIN are randomly inserted into the training data, then a language model is trained using word vectors as input features, and memorization is measured with a metric termed as exposure. The embedding dimension , the number of training epochs, and the length of the secret information were observed to affect memorization in pre-trained embeddings. Finally, to address the problem, differentially private language models were trained to reduce the exposure of sensitive information.
Document type :
Conference papers
Complete list of metadata

Cited literature [16 references]  Display  Hide  Download
Contributor : Zaineb Chelly Dagdia Connect in order to contact the contributor
Submitted on : Thursday, June 25, 2020 - 5:20:45 PM
Last modification on : Tuesday, August 4, 2020 - 11:10:02 AM
Long-term archiving on: : Wednesday, September 23, 2020 - 3:45:39 PM


Files produced by the author(s)


  • HAL Id : hal-02880590, version 1



Aleena Thomas, David Adelani, Ali Davody, Aditya Mogadala, Dietrich Klakow. Investigating the Impact of Pre-trained Word Embeddings on Memorization in Neural Networks. 23rd International Conference on Text, Speech and Dialogue, Sep 2020, brno, Czech Republic. ⟨hal-02880590⟩



Les métriques sont temporairement indisponibles