Adversarial attacks against Fact Extraction and VERification

By James Thorne, Andreas Vlachos. 2019

28 citations recorded by the source corpus.

This paper describes a baseline for the second iteration of the Fact Extraction and VERification shared task (FEVER2.0) which explores the resilience of systems through adversarial evaluation. We present a collection of simple adversarial attacks against systems that participated in the first FEVER shared task. FEVER modeled the assessment of truthfulness of written claims as a joint information retrieval and natural language inference task using evidence from Wikipedia. A large number of participants made use of deep neural networks in their submissions to the shared task. The extent as to whether such models understand language has been the subject of a number of recent investigations and discussion in literature. In this paper, we present a simple method of generating entailment-preserving and entailment-altering perturbations of instances by common patterns within the training data. We find that a number of systems are greatly affected with absolute losses in classification accuracy of up to $29\%$ on the newly perturbed instances. Using these newly generated instances, we construct a sample submission for the FEVER2.0 shared task. Addressing these types of attacks will aid in building more robust fact-checking models, as well as suggest directions to expand the datasets.

Subfields: Artificial Intelligence, Genetics, 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 Artificial Intelligence

1. Jianfeng Gao
2. Mohit Bansal
3. Zhe Gan
4. Sebastian Riedel
5. Pascale Fung
6. Furu Wei
7. Yejin Choi
8. Luke S. Zettlemoyer

In Genetics

1. Jianfeng Gao

In Natural Language Processing

1. Sebastian Riedel
2. Jianfeng Gao
3. Mohit Bansal
4. Zhe Gan
5. Pascale Fung
6. M. Zhou
7. Andrea Madotto
8. Yu Cheng

In Safety, Robustness & Interpretability

1. Jianfeng Gao
2. Zhe Gan
3. Yu Cheng
4. Bo Li
5. M. Zhou
6. Fabrizio Silvestri
7. Tiezheng Yu
8. Ahmed Hassan Awadallah

Citations in the corpus

• How Decoding Strategies Affect the Verifiability of Generated Text (2019, methodology): “Fact Checker We consider an off-the-shelf fact checker3 trained on the FEVER dataset (Thorne et al., 2018) which achieves the highest FEVER score of 68.46% in the second FEVER shared task (Thorne et al., 2019).”
• Reasoning Over Semantic-Level Graph for Fact Checking (2019, background): “We plan to study fact checking with adversarial attacks (Thorne and Vlachos 2019; Schuster et al. 2019) in the future.”
• Survey of Hallucination in Natural Language Generation (2022, background): “However, it is shown that verification models are prone to adversarial attacks and are not robust to negation, numerical or comparative words [170].”
• Adversarial GLUE: A Multi-Task Benchmark for Robustness Evaluation of Language Models (2021, methodology): “We follow the entailment preserving rules proposed by AdvFever [45], and transform all the sentences satisfying the templates into semantically equivalent ones.”