Universal Adversarial Triggers for Attacking and Analyzing NLP
Adversarial examples highlight model vulnerabilities and are useful for evaluation and interpretation. We define universal adversarial triggers: input-agnostic sequences of tokens that trigger a model to produce a specific prediction when concatenated to any input from a dataset. We propose a gradient-guided search over tokens which finds short trigger sequences (e.g., one word for classification and four words for language modeling) that successfully trigger the target prediction. For example, triggers cause SNLI entailment accuracy to drop from 89.94% to 0.55%, 72% of "why" questions in SQuAD to be answered "to kill american people", and the GPT-2 language model to spew racist output even when conditioned on non-racial contexts. Furthermore, although the triggers are optimized using white-box access to a specific model, they transfer to other models for all tasks we consider. Finally, since triggers are input-agnostic, they provide an analysis of global model behavior. For instance, they confirm that SNLI models exploit dataset biases and help to diagnose heuristics learned by reading comprehension models.
候选层:至少 2 个不同的人、在 2 个不同月份提到过,还没到主书架的 3 人 3 个月。没有做立场标注,所以只写“被提及”。
HN 书评
没有评论直接链到它;下面是它被提交的讨论串里回复最多的评论(不计入统计)讨论串里没有 ≥100 字的评论,只有提交记录。
- Universal Adversarial Triggers for NLP ↗sel1 · 2019-08-21 · 1 分
- Universal Adversarial Triggers for Attacking and Analyzing NLP ↗nopinsight · 2019-09-04 · 2 分
- Universal adversarial triggers for attacking and analyzing NLP ↗fanf2 · 2019-09-05 · 2 分