Kind of killed the fun, I normally don’t do reverse engineering challenges not my field. But I was interested how frontier models would handle this. Claude was of no help as opus 5.5 refused to do the work because of model constraints and 4.8 made a mess. Then switched to gpt Astra and it solved all 9 in 1h40minutes autonomously.
How do we feel about this?
[link] [comments]
from hacking: security in practice https://ift.tt/L48uvHZ
Comments
Post a Comment