Comment by phonon 7 days ago https://www.codereviewbench.com/https://codereview.withmartian.com/ 2 comments phonon Reply wordpad 6 days ago Based 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 5 days ago I will be trying these and will report back!
wordpad 6 days ago Based 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.
Based 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.
I will be trying these and will report back!