Show HN: I extracted the safety filters from Apple Intelligence models

2 BlueFalconHD 0 7/6/2025, 7:50:02 PM github.com ↗
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.

Comments (0)

No comments yet