Lune

EMNLP Findings2020

Assessing Robustness of Text Classification through Maximal Safe Radius Computation

Emanuele La Malfa, Min Wu, Luca Laurenti, Benjie Wang, Anthony Hartshorn, Marta Kwiatkowska

Abstract

Neural network NLP models are vulnerable to small modifications of the input that maintain the original meaning but result in a different prediction. In this paper, we focus on robustness of text classification against word substitutions, aiming to provide guarantees that the model prediction does not change if a word is replaced with a plausible alternative, such as a synonym. As a measure of robustness, we adopt the notion of the maximal safe radius for a given input text, which is the minimum distance in the embedding space to the decision boundary. Since computing the exact maximal safe radius is not feasible in practice, we instead approximate it by computing a lower and upper bound. For the upper bound computation, we employ Monte Carlo Tree Search in conjunction with syntactic filtering to analyse the effect of single and multiple word substitutions. The lower bound computation is achieved through an adaptation of the linear bounding techniques implemented in tools CNN-Cert and POPQORN, respectively for convolutional and recurrent network models. We evaluate the methods on sentiment analysis and news classification models for four datasets (IMDB, SST, AG News and NEWS) and a range of embeddings, and provide an analysis of robustness trends. We also apply our framework to interpretability analysis and compare it with LIME.

Ask Lune

Know this paper in minutes, not an afternoon.

Lune can read every page of it and the top-tier papers that cite it, and gives you its claims, its evidence and its limits, each one cited to the paper it comes from. No skimming, no AI slop.

1 top-tier paper has cited it since. Know what they found before your reviewers do.

Questions to start from

Lune can draw on

the full paper · the top-tier papers that cite it · related work

Get the answer

Free to start. No credit card required.

Lineage

  • The King Is Naked: On the Notion of Robustness for Natural Language Processing cites Seq2Sick: Evaluating the Robustness of Sequence-to-Sequence Models with Adversarial Examples
Builds onCited byLeft to right by year; the more cited, the higher.

Cited by top-tier papers1

  1. The King Is Naked: On the Notion of Robustness for Natural Language ProcessingEmanuele La Malfa, Marta KwiatkowskaAAAI 2022 · 31 citations

Builds on3

  1. Seq2Sick: Evaluating the Robustness of Sequence-to-Sequence Models with Adversarial ExamplesMinhao Cheng, Jinfeng Yi, Pin-Yu Chen, Huan Zhang et al.AAAI 2020 · 268 citations
  2. Robustness of Bayesian Neural Networks to Gradient-Based AttacksGinevra Carbone, Matthew Wicker, Luca Laurenti, Andrea Patané et al.NeurIPS 2020 · 85 citations
  3. Robustness Guarantees for Deep Neural Networks on VideosMin Wu, Marta KwiatkowskaCVPR 2020

Related papers

  1. TextVerifier: Robustness Verification for Textual Classifiers with Certifiable GuaranteesSiqi Sun, Wenjie RuanACL Findings 2023 · 2 citations
  2. Text-CRS: A Generalized Certified Robustness Framework against Textual Adversarial AttacksXinyu Zhang, Hanbin Hong, Yuan Hong, Peng Huang et al.S&P 2024 · 41 citations
  3. Robustness-Aware Word Embedding Improves Certified Robustness to Adversarial Word SubstitutionsYibin Wang, Yichen Yang, Di He, Kun HeACL Findings 2023 · 8 citations
  4. Towards Robustness Against Natural Language Word SubstitutionsXinshuai Dong, Anh Tuan Luu, Rongrong Ji, Hong LiuICLR 2021 · 63 citations
  5. Cert-RNN: Towards Certifying the Robustness of Recurrent Neural NetworksTianyu Du, Shouling Ji, Lujia Shen, Yao Zhang et al.CCS 2021 · 25 citations
Dusk over the sea between two cliffs drawn in fine vertical lines