Question

Assume a simple one-layer RNN whose computations depend on three parameters w, u, and v as follows: a linear state update function h_t = w cdot h_t-1 + u cdot x_t + v where h_t is the state at time step t, x_t is the input at time step t, and w, u, and v are the parameters. What is the function used for the linear state update in this RNN?

1) h_t = w cdot h_{t-1} + u cdot x_t + v
2) h_t = w cdot h_{t-1} + u cdot x_t
3) h_t = w cdot h_{t-1} + v
4) h_t = u cdot x_t + v

Answer 1

The function used for the linear state update in this RNN is h_t = w ⋅ h_t-1 + u ⋅ x_t + v. Option 1.

Step-by-step explanation:

The function used for the linear state update in this RNN is ht = w ⋅ ht-1 + u ⋅ xt + v. The linear state update function in the RNN is expressed as h_t = w * h_{t-1} + u * x_t + v, where h_t is the state at time step t, which incorporates the influence of the previous state, the current input, and parameters w, u, and v.

The function used for the linear state update in the described Recurrent Neural Network (RNN) is: h_t = w \cdot h_{t-1} + u \cdot x_t + v. This equation updates the state of the network at a given time step t by taking into account the previous state h_{t-1}, the current input x_t, and the parameters w, u, and v. The correct answer to the question is option 1.