Universal Adversarial Triggers for NLP

By Eric Wallace, Shi Feng, Nikhil Kandpal, Matt Gardner, Sameer Singh. 2019

30 citations recorded by the source corpus.

Subfields: Natural Language Processing, Safety, Robustness & Interpretability

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Natural Language Processing

In Safety, Robustness & Interpretability

Citations in the corpus

Plug and Play Language Models: A Simple Approach to Controlled Text Generation (2019, methodology): “Although the loss is a function of the entire sequence, here we adopt a greedy approach, similar to Ebrahimi et al. (2018); Wallace et al. (2019), in which we optimize for Table 5: Sentence samples in triplets, generated by {baseline GPT-2, PPLM-Discrim POSITIVE , PPLM-Discrim NEGATIVE },…”
Learning the Difference that Makes a Difference with Counterfactually-Augmented Data (2019, background): “For example, otherwise state-of-the-art models have been shown to be vulnerable to synthetic transformations such as distractor phrases [11, 27], to misclassify paraphrased task [10, 22] and to fail on template-based modifications [25].”
Explaining The Efficacy of Counterfactually-Augmented Data (2020, background): “As a result, models have been shown to be vulnerable to synthetic transformations, such as distractor phrases (Jia & Liang, 2017; Wallace et al., 2019), document paraphrases (Iyyer et al., 2018; Pfeiﬀer et al., 2019), and meaning preserving synthetic modiﬁcations (Ribeiro et al., 2018; Glockner et…”
Thieves on Sesame Street! Model Extraction of BERT-based APIs (2019, background): “While our classifier transfers well to two different nonsensical distributions, adaptive adversaries could gen- erate nonsensical queries that fool membership classifiers (Wallace et al., 2019).”
CoCon: A Self-Supervised Approach for Controlled Text Generation (2020, background): “Language model-based text generations, while fluent, are prone to producing inappropriate content [30] or be exploited for means like fake news [35].”
Hierarchical Reinforcement Learning for Open-Domain Dialog (2019, background): “Training on standard dialog datasets collected on-line or from movie scripts often leads to malicious, aggressive, biased, or offensive responses (Curry and Rieser 2018; He and Glass 2018; Henderson et al. 2018; Wallace et al. 2019).”
Robustness Gym: Unifying the NLP Evaluation Landscape (2021, background): “, 2020), Universal Adversarial Triggers (Wallace et al., 2019a), OpenAttack (Zeng et al.”