DeepAI AI Chat
Log In Sign Up

Disentangled Text Representation Learning with Information-Theoretic Perspective for Adversarial Robustness

by   Jiahao Zhao, et al.

Adversarial vulnerability remains a major obstacle to constructing reliable NLP systems. When imperceptible perturbations are added to raw input text, the performance of a deep learning model may drop dramatically under attacks. Recent work argues the adversarial vulnerability of the model is caused by the non-robust features in supervised training. Thus in this paper, we tackle the adversarial robustness challenge from the view of disentangled representation learning, which is able to explicitly disentangle robust and non-robust features in text. Specifically, inspired by the variation of information (VI) in information theory, we derive a disentangled learning objective composed of mutual information to represent both the semantic representativeness of latent embeddings and differentiation of robust and non-robust features. On the basis of this, we design a disentangled learning network to estimate these mutual information. Experiments on text classification and entailment tasks show that our method significantly outperforms the representative methods under adversarial attacks, indicating that discarding non-robust features is critical for improving adversarial robustness.


page 1

page 2

page 3

page 4


InfoBERT: Improving Robustness of Language Models from An Information Theoretic Perspective

Large-scale language models such as BERT have achieved state-of-the-art ...

Decoder-free Robustness Disentanglement without (Additional) Supervision

Adversarial Training (AT) is proposed to alleviate the adversarial vulne...

Robust Unsupervised Graph Representation Learning via Mutual Information Maximization

Recent studies have shown that GNNs are vulnerable to adversarial attack...

Derivation of Information-Theoretically Optimal Adversarial Attacks with Applications to Robust Machine Learning

We consider the theoretical problem of designing an optimal adversarial ...

Learning Bias-Invariant Representation by Cross-Sample Mutual Information Minimization

Deep learning algorithms mine knowledge from the training data and thus ...

InfoGAN: Interpretable Representation Learning by Information Maximizing Generative Adversarial Nets

This paper describes InfoGAN, an information-theoretic extension to the ...

Unsupervised Adversarially-Robust Representation Learning on Graphs

Recent works have demonstrated that deep learning on graphs is vulnerabl...