jm + turing   3

Decoding the Enigma with Recurrent Neural Networks
I am blown away by this -- given that Recurrent Neural Networks are Turing-complete, they can actually automate cryptanalysis given sufficient resources, at least to the degree of simulating the internal workings of the Enigma algorithm given plaintext, ciphertext and key:
The model needed to be very large to capture all the Enigma’s transformations. I had success with a single-celled LSTM model with 3000 hidden units. Training involved about a million steps of batched gradient descent: after a few days on a k40 GPU, I was getting 96-97% accuracy!
machine-learning  deep-learning  rnns  enigma  crypto  cryptanalysis  turing  history  gpus  gradient-descent 
27 days ago by jm
Accidentally Turing-Complete
slightly ruined by the inclusion of some "deliberately Turing-complete" systems
turing  computation  software  via:jwz  turing-complete  accidents  automatons 
march 2014 by jm

Copy this bookmark: