conftrace_

llm defense

3 papers

Explore in graph

Co-occurring keywords

adversarial attack (1621) large language model (13587) backdoor attack (385) jailbreaking attack (26) jailbreak attack (225) inference efficiency (248) prompt optimization (190) red teaming (61) prompt injection (101) transfer attack (21)

Papers

Summary the Savior: Harmful Keyword and Query-based Summarization for LLM Jailbreak Defense NAACL 2025
Jailbreaking with Universal Multi-Prompts NAACL 2025
Efficient but Vulnerable: Benchmarking and Defending LLM Batch Prompting Attack ACL 2025