← Back to papers
2021 IJCNLP IJCNLP 2021

Defending Pre-trained Language Models from Adversarial Word Substitution Without Performance Sacrifice

Authors

Rongzhou Bao , Jiayi Wang , Hai Zhao
Download PDF

Related papers

Second Order WinoBias (SoWinoBias) Test Set for Latent Gender Bias Detection in Coreference Resolution 2021
Flesch-Kincaid is Not a Text Simplification Evaluation Metric 2021
Semantic Similarity Based Evaluation for Abstractive News Summarization 2021
Figurative Language in Recognizing Textual Entailment 2021
Sequence Models for Computational Etymology of Borrowings 2021