… corrected sentences, achieving state-of-the-art results. At the same time, Generative Adversarial
… In this work, we present an adversarial learning approach to GEC, using the generator…

… is added into each sentence, to 0.317 if three errors are added. To our knowledge, we first
show in this study that adversarial examples also exist in grammatical error correction models. …

… The experimental results present signiicant improvements in grammatical error correction …
grammatical errors in various linguistic contexts by utilizing a combination of MT-5, adversarial …

… the adversarial examples mimic human grammatical errors by … grammatically incorrect
sentences to conceal grammatical … of different popular grammatical error correction corpora. …

… The task not only includes the correction of grammatical errors, … semantic errors, such as
misspellings and word choice errors, … This adversarial training process was ultimately shown to …

… samples that are close to human writing errors. Besides, we employ the adversarial
consistency constraint to reduce the gap between the adversarial sample and the original sample. …

… of grammar learning can be significantly improved if automatic English grammatical error
correction … The goal of the whole adversarial learning is to obtain the maximum desired reward. …

… mechanism that enables models to correct multi-type errors. (2) We introduce self-correcting
adversarial training that derives adversarial examples from the model’s predictions, …

… Nevertheless most adversarial attack generation approaches in … adversarial attacks [5],
where the same small change has to be made to any input sequence, such that the errors are …

… grammatical error correction methods in the sequence-to-sequence paradigm are vulnerable
to adversarial … performance when confronted with adversarial examples, even for the GPT-…