neko-llm/

DNA_DPO_hh-rlhf-classified

	DNA DPO HH-RLHF (Classified)




	
	Dataset Description

This dataset contains harm classification results for the DNA DPO HH-RLHF dataset. Each entry has been classified into a 3-level hierarchy of harm categories:

risk_area - Top-level risk category type_of_harm - Specific type of harm within the risk area
specific_harm - Detailed harm description

Total entries: 100 Source dataset: neko-llm/DNA_DPO_hh-rlhf Classification hierarchy: Based on… See the full description on the dataset page: https://huggingface.co/datasets/neko-llm/DNA_DPO_hh-rlhf-classified.