How can I access the gates of LSTM/GRU ? Are they exposed somehow ? If not, any alternative?
What do you mean by accessing them? Acquiring the weights?
Not weights. I meant value of the gates – forget/reset/update etc. ?
Specifically, the value after sigmoid is what it means.
I see. Not with the provided
nn.[GRU|RNN|LSTM](Cell) classes. But certainly doable if you write your own variant. A good reference is probably the Cell classes’ implementation. e.g.