Are Multilingual BERT models robust? A Case Study on Adversarial Attacks for Multilingual Question Answering
arXiv:2104.07646
Abstract
Recent approaches have exploited weaknesses in monolingual question answering (QA) models by adding adversarial statements to the passage. These attacks caused a reduction in state-of-the-art performance by almost 50%. In this paper, we are the first to explore and successfully attack a multilingual QA (MLQA) system pre-trained on multilingual BERT using several attack strategies for the adversarial statement reducing performance by as much as 85%. We show that the model gives priority to English and the language of the question regardless of the other languages in the QA pair. Further, we also show that adding our attack strategies during training helps alleviate the attacks.
References in corpus (5)
- Bidirectional Attention Flow for Machine Comprehension
- DRCD: a Chinese Machine Reading Comprehension Dataset
- Multilingual Extractive Reading Comprehension by Runtime Machine Translation
- DuReader: a Chinese Machine Reading Comprehension Dataset from Real-world Applications
- What do Models Learn from Question Answering Datasets?