$x_t$	= the input vector
$f_t$	= the activation vector of the forget gate at time t
$i_t$	= the activation vector of input gate at time t
$o_t$	= the activation vector of the output gate at time t
$c_t$	= the memory cell activation vector at time t
$h_t$	= the hidden state activation vector at time t
${\mathbf{W}}_{\mathbf{f}}$, ${\mathbf{W}}_{\mathbf{i}}$, and ${\mathbf{W}}_{\mathbf{o}}$	= the input kernels for the respective gates
${\mathbf{U}}_{\mathbf{f}}$, ${\mathbf{U}}_{\mathbf{i}}$, and ${\mathbf{U}}_{\mathbf{o}}$	= the recurrent kernels for the respective gates
$b_f$, $b_i$, $b_o$, and $b_c$	= the biases
$\sigma$	= the logistic sigmoid function
∘	= the matrix product
$tanh$	= the hyperbolic tangent activation function