This symbol represents the hidden state of a recurrent neural network.
This symbol represents the hidden state of a recurrent neural network. There is a recurrence relation between different hidden states. This allows the network to have an internal memory, and retain information over time.
The symbol, \(\mathbf{h}\), represents the hidden state of a recurrent neural network (RNN). It is also known as the internal state. The RNN's hidden state depends on previous states, which makes them recurrently related. This property allows the RNN to have an internal memory, and retain information over time. Having an internal memory is useful for tasks such as natural language processing (context) and time series prediction.