RNN

A neural network architecture designed for sequential data that passes a hidden state from one time step to the next, enabling processing of variable-length sequences. A Recurrent Neural Network (RNN) is a neural network in which connections between nodes form directed cycles, creating internal memory that allows the network to process sequences of arbitrary length.

RNNs process input one element at a time, updating a hidden state vector at each step that carries information from previous steps — enabling tasks like text generation, time-series forecasting, and speech recognition.

Recurrent Neural Network is a foundational concept in Model Architecture. A neural network architecture designed for sequential data that passes a hidden state from one time step to the next, enabling processing of variable-length sequences.