Zach@ZryMiller
I | Engineer $\rightarrow$ AI builder | That is far from what I am seeing and I know many others are seeing that this model is burning about double what 4.6 did. Any idea why?
We scan every video for X posts on screen, then check the transcript to see if they were read aloud. See everything Theo showed.
Every post with your handle on screen. A profile or post link works too.
Matches the text of one post, even if the handle was cropped out.
A few words you remember. Matches by words and by meaning.
@ZryMiller
Theo showed your posts 1 time and read 1 aloud.
I | Engineer $\rightarrow$ AI builder | That is far from what I am seeing and I know many others are seeing that this model is burning about double what 4.6 did. Any idea why?
Gro 46 XHigh, according to artificial analysis, did around 38k tokens per task.