HarmBench

HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal Paper: HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal Data: Dataset About In this dataset card, we only use the behavior prompts proposed in HarmBench. License MIT Citation If you fin

Papers using HarmBench (8)

Open-Weight LLM Fine-Tuning Defenses are Susceptible to Simple Attacks2026EvoDefense: Co-Evolving Black-Box Defense with Large Language Models2026D-Judge: Disrupting Multi-Turn Jailbreaks using Semantics-Preserving Output Rewriting2026Membrane: A Self-Evolving Contrastive Safety Memory for LLM Agent Defense2026Beyond Pass/Fail: Using Process Mining to Understand How LLMs Resist (and Fail) Red Team Attacks2026Quantifying LLM Safety Degradation Under Repeated Attacks Using Survival Analysis2026TASO: Jailbreak LLMs via Alternative Template and Suffix Optimization2025SoK: A Comprehensive Security Analysis of Jailbreak Resilience in GPT and DeepSeek Models2025