A Recurrent Neural Network (RNN) is a class of neural networks designed for processing sequential data, such as time series, speech, or text. Unlike feedforward networks, RNNs have loops in their architecture, allowing information to persist across time steps — making them ideal for tasks where context from earlier inputs is essential.

The key feature of an RNN is its hidden state, which acts as memory. At each time step, the network takes an input and the hidden state from the previous step to compute a new hidden state and an output. This allows the model to capture temporal dependencies in sequences, such as understanding the meaning of a word in context based on preceding words.

RNNs are used for:

• Language modeling and text generation,
• Speech recognition,
• Sequence classification,
• Time series forecasting.

However, standard RNNs struggle with long-range dependencies due to the vanishing/exploding gradient problem during training. To address this, more advanced variants were introduced:

• LSTM (Long Short-Term Memory) networks,
• GRU (Gated Recurrent Unit) networks.

These gated architectures improve the model's ability to retain and manage memory over longer sequences.