White-to-Black: Efficient Distillation of Black-Box Adversarial Attacks
By Yotam Gil, Yoav Chai, Or Gorodissky, Jonathan Berant. 2019
56 citations recorded by the source corpus.
Adversarial examples are important for understanding the behavior of neural models, and can improve their robustness through adversarial training. Recent work in natural language processing generated adversarial examples by assuming white-box access to the attacked model, and optimizing the input directly against it (Ebrahimi et al., 2018). In this work, we show that the knowledge implicit in the optimization procedure can be distilled into another more efficient neural network. We train a model to emulate the behavior of a white-box attack and show that it generalizes well across examples. Moreover, it reduces adversarial example generation time by 19x-39x. We also show that our approach transfers to a black-box setting, by attacking The Google Perspective API and exposing its vulnerability. Our attack flips the API-predicted label in 42% of the generated examples, while humans maintain high-accuracy in predicting the gold label.
Subfields: Artificial Intelligence, Signal Processing
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Artificial Intelligence
- Eric Wallace
- Dawn Song
- Mitchell Stern
- Luke S. Zettlemoyer
- Graham Neubig
- Quoc V. Le
- Jonathan Berant
- Yejin Choi
In Signal Processing
- Dawn Song
- Mitchell Stern
- Eric Wallace
- Engin Kirda
- Aloysius Ka-Lau Mok
- Simon Pak Ho Chung
- Christopher Krügel
Citations in the corpus
- Imitation Attacks and Defenses for Black-box Machine Translation Systems (2020, result): “We also follow similar practices from past published works that attack production systems (Papernot et al., 2017; Brendel et al., 2018; Ilyas et al., 2018; Gil et al., 2019).”