←back to thread

I extracted the safety filters from Apple Intelligence models

(github.com)
534 points BlueFalconHD | 1 comments | 06 Jul 25 19:50 UTC | HN request time: 0.212s | source

I managed to reverse engineer the encryption (refered to as “Obfuscation” in the framework) responsible for managing the safety filters of Apple Intelligence models. I have extracted them into a repository. I encourage you to take a look around.
Show context
trebligdivad ◴[06 Jul 25 20:56 UTC] No.44483981[source]▶
Some of the combinations are a bit weird, This one has lots of stuff avoiding death....together with a set ensuring all the Apple brands have the correct capitalisation. Priorities hey!

https://github.com/BlueFalconHD/apple_generative_model_safet...

replies(11): >>44483999 #>>44484073 #>>44484095 #>>44484410 #>>44484636 #>>44486072 #>>44487916 #>>44488185 #>>44488279 #>>44488362 #>>44488856 #
matsemann ◴[06 Jul 25 22:21 UTC] No.44484636[source]▶
So it blocks it from suggesting to "execute" a file or "pass on" some information.
replies(2): >>44484685 #>>44488029 #
1. extraduder_ire ◴[07 Jul 25 08:34 UTC] No.44488029[source]▶
Yahoo had this problem years ago when they rewrote emails to avoid the term "eval". (trying to filter dangerous javascript) Famously producing the word "medireview".