论文标题
同义替代攻击真的是同义词替代攻击吗？
Are Synonym Substitution Attacks Really Synonym Substitution Attacks?
论文作者
论文摘要
在本文中，我们探讨了以下问题：同义替代攻击是否真的同义替代替代攻击（SSA）？我们通过检查SSA如何替换原始句子中的单词来解决这个问题，并表明仍有未解决的障碍使当前的SSA会产生无效的对抗样本。我们揭示了四种广泛使用的单词替代方法会产生大量无效的替代单词，这些单词是不语法的，或者不保留原始句子的语义。接下来，我们表明SSA中用于检测无效单词替换的语义和语法约束在检测无效的对抗性样本方面高度不足。
In this paper, we explore the following question: Are synonym substitution attacks really synonym substitution attacks (SSAs)? We approach this question by examining how SSAs replace words in the original sentence and show that there are still unresolved obstacles that make current SSAs generate invalid adversarial samples. We reveal that four widely used word substitution methods generate a large fraction of invalid substitution words that are ungrammatical or do not preserve the original sentence's semantics. Next, we show that the semantic and grammatical constraints used in SSAs for detecting invalid word replacements are highly insufficient in detecting invalid adversarial samples.
