Tunable Efficient Unitary Neural Networks (EUNN)

Hi, I wonder if there’s been a PyTorch implementation of,

Tunable Efficient Unitary Neural Networks (EUNN)

It’s something that definitely seems to be a solid piece of work !

@smth this seems like something FAIR must have in house already? You, Yann LeCun and Martin Arjovsky have been working on this for a quite a while if I remember correctly?


There isn’t a PyTorch implementation of this publicly available as far as I know.


It’s in Tensorflow by one of the authors, Li Jing,


Compared to LSTM, at least this is mathematically interpretable !

Multi-layer bi-directional LSTM works great, but you can’t do any theory on it?

1 Like

Late to the party, but I will leave this here for anyone who bumps into this conversation.

The last few days I have been working on a pytorch implementation which can be found here:

The speed could probably be increased when PyTorch finally supports complex tensors :wink: