AI Dictionary › Fondamenti AI

Recurrent Neural Network (RNN)

Rete neurale ricorrente (RNN)

A recurrent neural network, or RNN, is an architecture designed to process sequences like text, audio, or time series, keeping a memory of what it has seen before. It processes elements one at a time and holds an internal state it updates at each step, carrying context along. It's like reading a sentence word by word while remembering earlier words to grasp the overall meaning: without memory, each word would be disconnected from the rest. This makes it naturally suited to data where order is crucial.

Definition

RNNs enabled translation, speech recognition, and sequential prediction, but struggle to recall information far back in time due to vanishing gradients. Variants like LSTMs eased the problem, but for long sequences they've largely been surpassed by transformers, which process the whole sequence in parallel and capture long-range relationships more effectively.

Related terms

• Long Short-Term Memory (LSTM)
• Convolutional Neural Network (CNN)
• Transformer
• Attention Mechanism
• Deep Learning

More in Fondamenti AI

• LLM (Large Language Model)
• Token
• Context Window
• Temperature
• AI Hallucination
• Fine-tuning
• RAG (Retrieval Augmented Generation)
• Embedding

Put it into practice

• Real scenario: apply this concept with an instant AI verdict.
• Daily gym: short drills to train dictionary concepts.
• Playoff game: challenge other players, elimination-style.

📱 Download the Android app (beta) iOS coming soon

Say what you mean. Get what you need.

Grace Certified, the AI coach that trains and certifies your prompt engineering, by Agora Intelligence.