I benchmarked 25 models against doing nothing. None of them won(github.com)
3 pointsby kaustubhspatilan hour ago1 comment