Masked Adversarial Generation for Neural Machine Translation
This addresses the challenge of efficient and effective adversarial attacks in machine translation, though it is incremental as it builds on existing gradient-based methods.
The paper tackles the problem of generating adversarial attacks for Neural Machine Translation by learning an adversarial generator based on a language model, resulting in improved robustness and faster performance compared to existing methods.
Attacking Neural Machine Translation models is an inherently combinatorial task on discrete sequences, solved with approximate heuristics. Most methods use the gradient to attack the model on each sample independently. Instead of mechanically applying the gradient, could we learn to produce meaningful adversarial attacks ? In contrast to existing approaches, we learn to attack a model by training an adversarial generator based on a language model. We propose the Masked Adversarial Generation (MAG) model, that learns to perturb the translation model throughout the training process. The experiments show that it improves the robustness of machine translation models, while being faster than competing methods.