logoalt Hacker News

ninja3925today at 6:03 PM2 repliesview on HN

Out of curiosity, how was it discovered? You would have to look for it to find this linear combination.


Replies

jdifftoday at 7:07 PM

Without the system prompt, asking its name results in it responding with the name of the model they're ripping from. That would certainly draw your eyes to the right places.

show 1 reply
Aurornistoday at 6:21 PM

Check the linked GitHub issue. They explain their process.

Scroll past the first issue to find it. It’s further down.