サインイン
インサイト - NLP Security
Syntactic Ghost: An Imperceptible General-purpose Backdoor Attacks on Pre-trained Language Models
신문의 핵심 메시지는 PLM에 대한 신문의 공격이 효과적이고 보이지 않는 트리거를 사용하여 다양한 다운스트림 작업에 위협을 제공한다는 것이다.
TrojFSP: Trojan Insertion in Few-shot Prompt Tuning
TrojFSP introduces a method to address security challenges in few-shot prompt tuning, achieving high attack success rates while maintaining clean data accuracy.
TrojFSP: Trojan Insertion in Few-shot Prompt Tuning
TrojFSP introduces a method to address security challenges in few-shot prompt tuning by balancing attack success rate and clean data accuracy.
