Recurrent neural network
A recurrent neural network is a neural network that processes ordered inputs while carrying and updating an internal state across successive steps.
Natural Language Processing in Action, Second Edition
A recurrent neural network is a neural network architecture that processes sequences while maintaining a state influenced by earlier inputs.
- How it is used
- Models variable-length sequences for classification, token labeling, prediction, and generation.
- How deep it goes
- Builds a recurrent step and training loop with linear layers, then explains shared weights and backpropagation through time.
- Examples
- Uses surname nationality classification and a WikiText2 word-level language model to illustrate sequence prediction.
- What you can do after
- Builds an RNN from scratch with PyTorch linear layers, trains it on character sequences, and inspects predictions and hidden states.
Read how Natural Language Processing in Action, Second Edition explains recurrent neural network →
Sutskever's List
A recurrent neural network is a neural network architecture that processes sequences by maintaining a state updated across successive inputs.
- How it is used
- Models ordered text, speech, and time series by carrying context forward for later predictions.
- How deep it goes
- Explains hidden state, unrolled training, vanishing gradients, LSTM remedies, and translation bottlenecks without a formal derivation.
- Examples
- Uses a distant clue in a sentence, character-level essay generation, speech recognition, and machine translation.
- What you can do after
- Evaluate/understand when recurrent state helps sequence prediction and why long dependencies, overfitting, or fixed-length encoding can limit it.
Read how Sutskever's List explains recurrent neural network →
Deep Learning with Python, Third Edition
A neural network architecture that processes sequences by feeding information from one step to the next.
- How it is used
- Models ordered sequences for weather forecasting, character prediction, and translation.
- How deep it goes
- Explains state updates through a NumPy forward pass and develops Keras recurrent variants and encoder-decoder models.
- Examples
- Uses Jena weather forecasting, a Shakespeare character model, and English-to-Spanish translation.
- What you can do after
- Apply it via Keras to build recurrent forecasting, language modeling, and sequence-to-sequence models.
Read how Deep Learning with Python, Third Edition explains recurrent neural network →
Deep Learning with R, Third Edition
A recurrent neural network is a neural-network model that processes sequences recurrently by carrying information through an internal state across positions.
- How it is used
- Uses recurrent state for ordered data, especially forecasting, language modeling, and sequence-to-sequence translation.
- How deep it goes
- Works through a toy forward pass, recurrent-layer shapes, LSTM architecture, and encoder–decoder design.
- Examples
- Uses Jena weather forecasts, a Shakespeare character model, and English-to-Spanish translation.
- What you can do after
- Apply it via R and Keras to forecasting, character prediction, and sequence-to-sequence modeling.
Read how Deep Learning with R, Third Edition explains recurrent neural network →
Inside Deep Learning: Math, Algorithms, Models
A neural network that processes sequential data by reusing parameters across time steps and carrying a hidden state forward.
- How it is used
- Processes ordered, variable-length data and supplies a final sequence representation for classification.
- How deep it goes
- Works through recurrence, tensor shapes, final-state extraction, packed batching, and stacked layers; defers more advanced recurrent variants.
- Examples
- Builds a character-sequence classifier for a name's likely language and compares single-item with packed-batch training.
- What you can do after
- Apply it via PyTorch nn.RNN with embeddings, final-state extraction, a linear classifier, and packed variable-length batches.
Read how Inside Deep Learning: Math, Algorithms, Models explains recurrent neural network →
Graph Neural Networks in Action
A recurrent neural network is a neural network architecture that processes sequences using recurrent state updates.
- How it is used
- Provides a temporal baseline for predicting future body-joint sensor positions from past frames without edge data.
- How deep it goes
- Explains hidden-state updates and backpropagation through time, then implements an nn.RNN; leaves difficult gradient computations to deep-learning packages.
- Examples
- Uses body-joint sensor trajectories, an unfolded recurrent-state diagram, and PyTorch model, training, and forecasting code.
- What you can do after
- Apply it via PyTorch nn.RNN to forecast sensor positions and recognize when missing relational information limits predictions.
Read how Graph Neural Networks in Action explains recurrent neural network →
Learn Generative AI with PyTorch
A neural network architecture for sequential data that processes inputs with loops so information can persist across time steps.
- How it is used
- Uses RNNs to model ordered sequences and generate context-sensitive text.
- How deep it goes
- Explains hidden-state updates and limitations, then builds and trains an LSTM for text generation.
- Examples
- Illustrates next-word prediction with “a frog has four legs” and word-order sensitivity with a football result.
- What you can do after
- Build and train an LSTM for text generation, then adjust output diversity with temperature and top-K sampling.
Read how Learn Generative AI with PyTorch explains recurrent neural network →
Build AI Drug Discovery Pipelines
A recurrent neural network cell is the repeated computational unit that combines the current input with the previous hidden state to produce an output and updated hidden state.
- How it is used
- Processes SMILES as ordered sequences for molecular encoding and step-by-step generation.
- How deep it goes
- Explains hidden-state recurrence and a simple cell implementation, then addresses vanishing gradients with GRU gates.
- Examples
- Uses acetic acid, caffeine, benzene ring closures, and a PyTorch-style single-cell computation to illustrate recurrence.
- What you can do after
- Build it from scratch by implementing one recurrent step and selecting GRUs when molecular sequences require longer-range memory.
Read how Build AI Drug Discovery Pipelines explains recurrent neural network →
Build an Advanced RAG Application (From Scratch)
A recurrent neural network is a neural model that processes sequence elements iteratively while maintaining a state representing prior inputs.
- How it is used
- Frames recurrent neural networks as earlier sequence-modeling architectures against which Transformer processing is evaluated.
- How deep it goes
- Explains sequential hidden-state processing and training limitations, but stops short of parameter updates, equations, or implementation details.
- Examples
- Uses the sentence “The quick brown fox jumps over the lazy dog” to contrast sequential recurrence with Transformer processing.
- What you can do after
- Evaluate/understand why sequential recurrent models struggle with parallelization, distant dependencies, and unstable gradients.
Read how Build an Advanced RAG Application (From Scratch) explains recurrent neural network →
Fight Fraud with Machine Learning
A recurrent neural network is a neural-network model that processes ordered sequences by maintaining an internal state and using connections across time steps.
- How it is used
- Targets time series, text, and fraud-detection sequences where temporal dependencies affect predictions.
- How deep it goes
- Explains sequences, recurrent cycles, and hidden state, but stops short of detailed RNN variants or training mechanics.
- Examples
- Uses a dog-versus-cat pixel sequence and voice authentication for fraud detection.
- What you can do after
- Evaluate/understand when recurrent connections and hidden state suit sequential fraud-detection data.
Read how Fight Fraud with Machine Learning explains recurrent neural network →
Timeless Algorithms, The Foundational Ideas
A neural network that processes sequential inputs using a state carried across sequence positions.
- How it is used
- Processes ordered language, speech, and time-series inputs one step at a time, carrying prior context into later predictions.
- How deep it goes
- Explains hidden-state updates, fixed-size compression, long-range information loss, and sequential computation, while stopping short of implementation details.
- Examples
- Uses a nested-clause sentence to illustrate how a distant subject–verb relationship can be lost.
- What you can do after
- Evaluate recurrent models against transformers by tracing how context propagation affects distant dependencies and parallel training.
Read how Timeless Algorithms, The Foundational Ideas explains recurrent neural network →
Building Specialized AI Systems
A recurrent neural network is a sequence model that processes inputs sequentially while carrying a hidden state between positions.
- How it is used
- Uses sequential hidden-state processing to model temporal dependencies in NLP and other ordered data.
- How deep it goes
- Explains hidden-state updates and encoder-decoder compression, then stops at limitations without detailing training algorithms or cell equations.
- Examples
- Uses language translation to demonstrate token-by-token encoding, fixed-size memory, and decoder generation.
- What you can do after
- Evaluate sequential sequence models against transformers using parallelism, long-range memory, and representation-bottleneck trade-offs.
Read how Building Specialized AI Systems explains recurrent neural network →
Transfer Learning for Natural Language Processing
A recurrent neural network is a neural network that processes sequential data while maintaining state across sequence positions.
- How it is used
- Uses RNNs to model ordered text sequences and as components in NLP transfer-learning methods.
- How deep it goes
- Explains sequence handling and order encoding, but stops short of recurrent computations or training internals.
- Examples
- Uses translation, summarization, column-type classification, and fake-news detection as NLP applications.
- What you can do after
- Understand how recurrent encoders represent sequence order and where RNN-based transfer methods are applied.
Read how Transfer Learning for Natural Language Processing explains recurrent neural network →
LLMs in Production
A recurrent neural network is a neural network that processes ordered inputs while maintaining a state that carries information across sequence positions.
- How it is used
- Processes ordered language inputs while preserving information across sequence positions for language modeling and sentiment analysis.
- How deep it goes
- Explains state, variable-length inputs, and gradient failure modes, then points to LSTM remedies without deriving recurrent internals.
- Examples
- Uses paired sentiment sentences and a tokenized Twitter dataset with sentiment labels to illustrate sequence modeling.
- What you can do after
- Evaluate sequence-model trade-offs, especially variable-length support and gradient instability, before choosing an RNN or LSTM.
Read how LLMs in Production explains recurrent neural network →
Understanding AI
A recurrent neural network is a neural network that processes sequential inputs step by step while maintaining a state that carries information across the sequence.
- How it is used
- Frames language generation as sequential, context-based next-word prediction learned from data.
- How deep it goes
- Explains sequential processing, hidden state, embeddings, and long-context limits, but defers mathematical or implementation detail.
- Examples
- Uses the umbrella sentence to illustrate context accumulation and contextually appropriate continuation.
- What you can do after
- Evaluate why sequential models can predict coherently yet lose important information across longer inputs.
Read how Understanding AI explains recurrent neural network →
Grokking Machine Learning
A recurrent neural network is a neural network architecture that processes sequences by retaining and updating state across successive inputs.
- How it is used
- Processes text and other sequential inputs for language-understanding applications.
- How deep it goes
- Defines the architecture and its language-processing role but stops short of explaining its internal equations or training procedure.
- Examples
- Uses sentiment analysis, machine translation, speech recognition, text summarization, and chatbots as application examples.
- What you can do after
- Evaluate understanding of sequence-processing applications such as sentiment analysis, translation, and speech recognition.
Read how Grokking Machine Learning explains recurrent neural network →
Data Without Labels
A recurrent neural network is a neural-network architecture designed to model sequences through recurrent state or connections.
- How it is used
- Handles ordered datasets whose temporal dependencies influence pattern recognition.
- How deep it goes
- Defines recurrent state, shared weights, outputs, and gradient limits, but defers detailed LSTM and gated-unit treatment.
- Examples
- Uses time-series data, videos, natural language, speech recognition, and other ordered datasets as application contexts.
- What you can do after
- Evaluate sequence-modeling choices by relating order, recurrent memory, and long-term dependency limits to data requirements.
Read how Data Without Labels explains recurrent neural network →
Machines that Think
A recurrent architecture processes sequences by carrying information from earlier sequence elements into later computations.
- How it is used
- Processes ordered data by carrying context from one step to the next, supporting sequence tasks such as prediction and tagging.
- How deep it goes
- Explains hidden-state updates and shared weights, then identifies long-range memory limits without detailing training procedures.
- Examples
- Uses words, music, sensor readings, next-word prediction, tagging, and a sentence whose subject spans intervening phrases.
- What you can do after
- Understand how a hidden state carries sequence context and why compression can weaken or entangle earlier information.
Read how Machines that Think explains recurrent neural network →
Also covered in
- Grokking AI Algorithms, Second Edition
- Classic Computer Science Problems in Swift: Essential techniques for practicing programmers
- Transformers in Action
- Domain-Specific Small Language Models
- Build a Large Language Model (From Scratch)
- Machine Learning Algorithms in Depth
- Build AI-Enhanced Web Apps
- Grokking Machine Learning, Second EditionIn early access (MEAP)
What to know first
- language modelA language model is a computational model that represents patterns in linguistic sequences to assign probabilities, predict, generate, or evaluate language.
- LSTMA long short-term memory (LSTM) network is a recurrent neural-network architecture with gated memory cells that control information flow and model dependencies in sequential data.
- tokenA token is a discrete unit of text, such as a word, subword, punctuation mark, or symbol, produced by tokenization and processed by a language model.
- accuracyAccuracy is the degree to which a result agrees with a specified truth, reference, or criterion, commonly measured as the proportion of evaluated predictions that are correct.
Related concepts
- LSTMincludes · A long short-term memory (LSTM) network is a recurrent neural-network architecture with gated memory cells that control information flow and model dependencies in sequential data.
- sequenceuses · A sequence is an ordered collection of elements whose positions determine the order in which they are arranged or traversed.
- language modelkind of · A language model is a computational model that represents patterns in linguistic sequences to assign probabilities, predict, generate, or evaluate language.
- hidden stateuses · An internal condition or representation of a system that is not directly observable but can influence its outputs or subsequent processing.
- bidirectional rnnused by · A bidirectional recurrent neural network processes a sequence in forward and reverse directions and combines the resulting representations.
- statehas part · State is information describing the condition or configuration of a system or computation at a particular time that may influence its behavior or subsequent transitions.
Questions about recurrent neural network
What is recurrent neural network?
A recurrent neural network is a neural network that processes ordered inputs while carrying and updating an internal state across successive steps.
Which Manning books explain recurrent neural network?
Natural Language Processing in Action, Second Edition, Sutskever's List, Deep Learning with Python, Third Edition, Deep Learning with R, Third Edition, Inside Deep Learning: Math, Algorithms, Models, Graph Neural Networks in Action, Learn Generative AI with PyTorch, Build AI Drug Discovery Pipelines and 10 more.
What should I know before learning recurrent neural network?
Manning books that teach recurrent neural network usually introduce language model, LSTM, token and accuracy first.
What kinds of recurrent neural network are there?
Kinds of recurrent neural network include LSTM.
What is recurrent neural network a kind of?
Recurrent neural network is a kind of language model. A language model is a computational model that represents patterns in linguistic sequences to assign probabilities, predict, generate, or evaluate language.
What are the parts of recurrent neural network?
Recurrent neural network includes state.
These guides are generated from the text of Manning books by Manning's knowledge graph and link to the sections they draw on. The full books are available on liveBook and at manning.com.

















