This paper teaches a new way to make a language model pay extra attention to the exact words you highlight in a prompt.
NanoKnow is a new benchmark that checks whether a language model’s answers come from what it saw during training or from extra text we give it at question time.