Data as of Jul 25, 2026 · Based on 2,718,867 AI responses across 9,511 prompts · See how Parse measures this
RealToxicityPromptshuggingface.coThe Real Toxicity Prompts dataset on Hugging Face provides text prompts with toxicity annotations for training and evaluating language models. It includes over 100,000 examples with human-validated toxicity scores across multiple categories.
Parse Score
