It's clear how LSTM works with 1 feature. But what happens if the number of features is > 1?
According to the answer proposed here,
Keras creates a computational graph that executes the sequence in your bottom picture per feature (but for all units). That means the state value C is always a scalar, one per unit.
But if it executed the process per feature, the result would be
number of units x number of features. Instead, we get only the number of units (if
number of timesteps x number of units).
What happens to features and how are they processed? At which step and how are they merged?