Gated update
A gated update is a state or representation update in which learned gates control how new information is combined with or replaces existing information.
Build AI Drug Discovery Pipelines
An update gate is a gated-recurrent-unit mechanism that controls how much of a candidate state replaces or supplements the previous hidden state.
- How it is used
- Controls memory updates in GRUs processing long, complex molecular sequences.
- How deep it goes
- Defines the gate, gives its blending equation, and connects it to GRU cell implementation.
- Examples
- Uses SMILES parsing, matching parentheses, and ring closures to illustrate retained sequence information.
- What you can do after
- Build GRU update behavior by implementing its gate and hidden-state blend.
Read how Build AI Drug Discovery Pipelines explains gated update →
Natural Language Processing in Action, Second Edition
An update gate is a gated recurrent mechanism that controls how much of the previous hidden state is retained versus replaced by new information.
- How it is used
- Controls recurrent memory flow so GRU units can handle information that remains relevant across tokens.
- How deep it goes
- Explains gate roles and gives the sigmoid computation, but stops short of deriving the complete hidden-state update.
- Examples
- Uses GRU pseudocode computing reset, update, and candidate gates from input and hidden-state matrix products.
- What you can do after
- Understand how sigmoid gate values regulate recurrent hidden-state information at each token.
Read how Natural Language Processing in Action, Second Edition explains gated update →
Also covered in
What to know first
- backpropagationBackpropagation is reverse-mode differentiation that propagates derivatives backward through a computational graph to compute gradients of an output or loss with respect to inputs and parameters.
- recurrent neural networkA recurrent neural network is a neural network that processes ordered inputs while carrying and updating an internal state across successive steps.
Related concepts
- GRUpart of · A gated recurrent unit (GRU) is a recurrent neural-network architecture that uses learned gates to regulate information flow and update a hidden state across sequence steps.
- jumping knowledge networkalternative to · A jumping knowledge network is a graph neural-network architecture that combines node representations from multiple message-passing depths for final predictions.
- reset gatecontrasts with · A reset gate is a GRU mechanism that controls how much of the previous hidden state contributes to computing a candidate state.
Questions about gated update
What is gated update?
A gated update is a state or representation update in which learned gates control how new information is combined with or replaces existing information.
Which Manning books explain gated update?
Build AI Drug Discovery Pipelines and Natural Language Processing in Action, Second Edition.
What should I know before learning gated update?
Manning books that teach gated update usually introduce backpropagation and recurrent neural network first.
What is the difference between gated update and jumping knowledge network?
A gated update is a state or representation update in which learned gates control how new information is combined with or replaces existing information. A jumping knowledge network is a graph neural-network architecture that combines node representations from multiple message-passing depths for final predictions.
What is gated update a kind of?
Gated update is a kind of update strategy. An update strategy is a method for combining a node's previous representation with information aggregated from its neighborhood.
What is gated update part of?
Gated update is part of GRU. A gated recurrent unit (GRU) is a recurrent neural-network architecture that uses learned gates to regulate information flow and update a hidden state across sequence steps.
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.

