The repositort repeats the paper: "Neural Cleanse: Identifying and Mitigating Backdoor Attacks in Neural Networks".