Factual Accuracy and Reliability Benchmarks
#29
by
Taylor658
- opened
Even though Gemma's primary task's are question answering, summarization, and reasoning and not explicit fact checking; will factual accuracy and reliability benchmarks be added in the future to the Model Card? I am primarily interested in FEVER and Climate-FEVER benchmarks for fact checking accuracy.
Thanks for the feedback Taylor. I'll pass this to the research team!