Skip to content Skip to sidebar Skip to footer
Search
Search
Search

An improved, more efficient method to prohibit an AI chatbot from producing harmful responses.

Researchers from Improbable AI Lab at MIT and the MIT-IBM Watson AI Lab…