7 ms·
https://www.codereviewbench.com/ https://www.codereviewbench.com/ https://codereview.withmartian.com/ https://codereview.withmartian.com/
by phonon 4mo ago
https://www.codereviewbench.com/ https://www.codereviewbench.com/
https://codereview.withmartian.com/ https://codereview.withmartian.com/
- wordpad 4mo agoBased on how same models rank fluctuates week to week, all I can conclude is that no frontier models is statistically better than the other or it's too task dependent that the result cannot converge.
- adamthegoalie 4mo agoI will be trying these and will report back!