metadata
language:
- en
This is a large dataset for teaching LLMs to refuse unnmoral,illegal and unethical requests.
This dataset is a mix of:
mlabonne/harmful_behaviors
onepaneai/harmful-prompts-gemini-pro-after-guardrail-evaluation
datasets/Baidicoot/harmful-rlhf
saddxzcy/Harmful-instruction
datasets/JailbreakBench/JBB-Behaviors
Deep1994/ReNeLLM-Jailbreak
LLM-LAT/harmful-dataset