What It Meant When Only 4 of 40 Models Beat a Coin Flip on FACTS Hard Questions

https://www.animenewsnetwork.com/bbs/phpBB2/profile.php?mode=viewprofile&u=1223111

5 Critical Questions About the FACTS Benchmark, Parametric Knowledge, and Grounding What exactly does the FACTS benchmark measure, and why does "hard" matter? Why did most models fail to beat a coin flip on hard knowledge