News for hackers. Stuff that matters.
Re: by markasoftware on Saturday October 12, 2024 @07:23PM (#41823638)
You can predict the first bit of sha256 with 100% accuracy with the following pseudocode:

`sha256sum(input) & 0x1`

What's the significance of being able to predict that bit with an ML model instead?

Re: by seccode on Saturday October 12, 2024 @06:22PM (#41823235)
Would greatly appreciate any peer review for this work
Re: by vednig on Saturday October 12, 2024 @07:18PM (#41823607)
I hv starred the repo will come back to you in few hours
Re: by xtrapol8 on Saturday October 12, 2024 @06:30PM (#41823285)
Oh no! I was hoping sha256 would last us until the quantum computing cow came home.
Re: by seccode on Saturday October 12, 2024 @06:32PM (#41823301)
This doesn't _break_ sha256, just opens the door to breaking sha256 with machine learning
Re: by Havoc on Saturday October 12, 2024 @07:22PM (#41823622)
Just is did NSA and friends already slip through the door
Re: by big-green-man on Saturday October 12, 2024 @06:43PM (#41823372)
Dude you're on fire the last couple of days with the random forest classifier.