← emma-w1's Projects
CALM-RAG500
BLUFF-1000 is a comprehensive benchmark that evaluates the factuality, faithfulness, and uncertainty expression abilities of RAG models. It contains 500 questions and 1000 evaluation instances.
CALM-RAG500
by @emma-w1BLUFF-1000 is a comprehensive benchmark that evaluates the factuality, faithfulness, and uncertainty expression abilities of RAG models. It contains 500 questions and 1000 evaluation instances.
12.45 hours