RMLM: A Flexible Defense Framework for Proactively Mitigating Word-level Adversarial Attacks

Zhaoyang Wang, Zhiyue Liu, Xiaopeng Zheng, Qinliang Su, Jiahai Wang