logoalt Hacker News

irthomasthomastoday at 8:55 PM1 replyview on HN

Or they trained a LoRA on the victims chats in order to launder their plagiarism.


Replies

fertoday at 9:19 PM

The timing makes it the most likely, not only them but potentially more. Comparatively quick, instant results. "Here's Astra! BTW our internal model is 10x better at math!" It'd be interesting to see academics having giving deeper looks at whatever OpenAI publishes from now on.