Break the Breakout: Reinventing LM Defense Against Jailbreak Attacks with Self-Refinement Paper β’ 2402.15180 β’ Published Feb 23
FLEX: Expert-level False-Less EXecution Metric for Reliable Text-to-SQL Benchmark Paper β’ 2409.19014 β’ Published Sep 24
Korean LLM Safety Guard Collection μμ ν νκ΅μ΄ μΈμ΄λͺ¨λΈμ νμ΅ λ° νκ°λ₯Ό μν λꡬλ€μ λͺ¨μμ΅λλ€. λ²μλ νμ΅ λ° νκ° λ°μ΄ν°μ μ΄λ₯Ό κΈ°λ°μΌλ‘ νμ΅ν Guard λͺ¨λΈμ ν¬ν¨νκ³ μμ΅λλ€. β’ 5 items β’ Updated Oct 25