Showing cs.CRShow all
2 papers · 1 filter
cs.CR2026
Why Are LLM Backdoor Defenses Fragmented? A Feature-Level Explanation with Sparse Autoencoders
Yizhe Zeng, Chenxu Niu, Wei Zhang +7
Backdoor attacks pose a serious threat to large language models (LLMs), but existing defenses remain fragmented, failing to pro?vide unified defense against both dirty-label and cl…
cs.CR2026
MirageBackdoor: A Stealthy Attack that Induces Think-Well-Answer-Wrong Reasoning
Yizhe Zeng, Wei Zhang, Yunpeng Li +3
While Chain-of-Thought (CoT) prompting has become a standard paradigm for eliciting complex reasoning capabilities in Large Language Models, it inadvertently exposes a new attack s…