undefined | Better HN

0 pointsadastra226mo ago0 comments

That is not how training works…

0 comments

That's how model distillation works.

DeepSeek is the most notable case, but it's been used lots.

And the foundation model companies are scraping and exfiltrating each others' data.

j / k navigate · click thread line to collapse

0 pointsadastra226mo ago0 comments

That is not how training works…

That's how model distillation works.

DeepSeek is the most notable case, but it's been used lots.

And the foundation model companies are scraping and exfiltrating each others' data.

j / k navigate · click thread line to collapse