MLCommons releases new benchmark for measuring LLM safety – works by supplying an LLM with over 24,000 prompts and harmful content created for safety evaluation purposes December 5, 2024 // by Finnovate This content is for members only. Sign up for access to the latest trends and innovations in fintech. View subscription plans. Login