How do I get that loss, though, without the softmax inputs? | Hacker News

Hacker Newsnew | past | comments | ask | show | jobs | submit

CamperBob2 86 days ago | parent | context | favorite | on: The text in Claude Code’s “Extended Thinking” outp...

How do I get that loss, though, without the softmax inputs?

wren6991 86 days ago [–]

Do they have logits for all of the Wikipedia etc that they've scraped?

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact