cherry opel project
It is a classification of prompts as safe or harmful, and not only that, but also determining the type of harm present in the prompt, then determining the text that contains harm, comparing two or more responses, and working to reach the best image for the artificial intelligence response.