Adam optimizer
E182821
The Adam optimizer is a popular stochastic gradient descent method in machine learning that adaptively adjusts learning rates for each parameter using estimates of first and second moments of gradients.
All labels observed (4)
| Label | Occurrences |
|---|---|
| Adam optimizer canonical | 5 |
| Adam optimization algorithm | 2 |
| AdamW | 1 |
| AdamW optimizer | 1 |
How this entity was disambiguated
This entity first appeared as the object of triple T1616490 — resolving that mention is where its identity was fixed. The disambiguator weighed these candidate entities and picked the highlighted one (or “None”, minting a new entity). This is how homonymy is resolved: the same surface form can point to different entities.
NED1
Entity disambiguation (via context triple)
gpt-5-mini-2025-08-07
Target entity: Adam optimizer Context triple: [Jimmy Ba, knownFor, Adam optimizer]
-
A.
RMSProp
RMSProp is an adaptive gradient-based optimization algorithm commonly used to efficiently train deep neural networks by adjusting learning rates for individual parameters.
-
B.
Automatic Adam
Automatic Adam is the nickname of Adam Vinatieri, a legendary NFL placekicker renowned for his clutch, game-winning field goals in high-pressure situations.
-
C.
TRPO
TRPO (Trust Region Policy Optimization) is a reinforcement learning algorithm that optimizes policies with guaranteed monotonic improvement by constraining each update within a trust region to maintain stability.
-
D.
SGD
SGD is the official currency code for the Singapore dollar, the national currency of Singapore used in domestic and international transactions.
-
E.
DDPG
DDPG (Deep Deterministic Policy Gradient) is a model-free, off-policy deep reinforcement learning algorithm designed for continuous action spaces, combining ideas from DQN and actor-critic methods.
- F. None of above. chosen
- G. Unsure - the case is ambiguous/there is not enough information to decide.
NED2
Entity disambiguation (via description)
gpt-5-mini-2025-08-07
Target entity: Adam optimizer Target entity description: The Adam optimizer is a popular stochastic gradient descent method in machine learning that adaptively adjusts learning rates for each parameter using estimates of first and second moments of gradients.
-
A.
RMSProp
RMSProp is an adaptive gradient-based optimization algorithm commonly used to efficiently train deep neural networks by adjusting learning rates for individual parameters.
-
B.
Automatic Adam
Automatic Adam is the nickname of Adam Vinatieri, a legendary NFL placekicker renowned for his clutch, game-winning field goals in high-pressure situations.
-
C.
TRPO
TRPO (Trust Region Policy Optimization) is a reinforcement learning algorithm that optimizes policies with guaranteed monotonic improvement by constraining each update within a trust region to maintain stability.
-
D.
SGD
SGD is the official currency code for the Singapore dollar, the national currency of Singapore used in domestic and international transactions.
-
E.
DDPG
DDPG (Deep Deterministic Policy Gradient) is a model-free, off-policy deep reinforcement learning algorithm designed for continuous action spaces, combining ideas from DQN and actor-critic methods.
- F. None of above. chosen
Statements (56)
| Predicate | Object |
|---|---|
| instanceOf |
adaptive learning rate method
ⓘ
optimization algorithm ⓘ stochastic gradient descent variant ⓘ |
| baseAlgorithm | stochastic gradient descent ⓘ |
| category | adaptive gradient method ⓘ |
| commonlyUsedIn |
computer vision models
ⓘ
deep learning ⓘ natural language processing models ⓘ neural network training ⓘ reinforcement learning ⓘ |
| defaultBeta1 | 0.9 ⓘ |
| defaultBeta2 | 0.999 ⓘ |
| defaultEpsilon | 1e-8 ⓘ |
| defaultLearningRate | 0.001 ⓘ |
| fullName |
Adam: A Method for Stochastic Optimization
ⓘ
surface form:
Adaptive Moment Estimation
|
| gradientRequirement | first-order gradients ⓘ |
| hasVariant |
AdaGrad
ⓘ
surface form:
AMSGrad
Adam optimizer self-linksurface differs ⓘ
surface form:
AdamW
|
| hyperparameter |
beta1
ⓘ
beta2 ⓘ epsilon ⓘ learning rate ⓘ weight decay (in some implementations) ⓘ |
| implementedIn |
Chainer
ⓘ
JAX ⓘ Keras ⓘ MXNet ⓘ PyTorch ⓘ TensorFlow ⓘ fastai ⓘ |
| inspiredBy |
AdaGrad
ⓘ
RMSProp ⓘ |
| introducedBy |
Diederik P. Kingma
ⓘ
Jimmy Ba ⓘ |
| introducedInPaper | Adam: A Method for Stochastic Optimization ⓘ |
| maintains |
per-parameter first moment vector
ⓘ
per-parameter second moment vector ⓘ |
| optimizationTarget | minimization of loss function ⓘ |
| parameterUpdateDependsOn |
current gradient
ⓘ
first moment estimate ⓘ second moment estimate ⓘ |
| performsBiasCorrection | true ⓘ |
| publicationYear | 2014 ⓘ |
| relatedTo |
AdaDelta
ⓘ
AdaGrad ⓘ RMSProp ⓘ |
| strength | fast convergence in practice ⓘ |
| suitableFor |
large-scale problems
ⓘ
non-stationary objectives ⓘ sparse gradients ⓘ |
| updateRuleType | adaptive learning rate ⓘ |
| uses |
exponentially decaying averages of past gradients
ⓘ
exponentially decaying averages of past squared gradients ⓘ |
| usesFirstMomentEstimate | gradient mean ⓘ |
| usesSecondMomentEstimate | uncentered gradient variance ⓘ |
| weakness | can generalize worse than SGD with momentum in some settings ⓘ |
How these facts were elicited
The pipeline generated the facts above by prompting gpt-5.1 with this entity's name + description and the instruction below.
Instruction
You are a knowledge base construction expert. Given a subject entity and a description of it, return factual statements that you know for the subject as a JSON list of dictionaries(triples), where keys must be "subject", "predicate" and "object". The number of facts may be very high, between 25 to 50 or more, for very popular subjects. For less popular subjects, the number of facts can be very low, like 5 or 10. # Requirements - If you don't know the subject at all, return an empty list. - If the subject is not a named entity, return an empty list. - Include at least one triple where predicate is "instanceOf". - Do not get too wordy. - Separate several objects into multiple triples with one object.
Input
Subject: Adam optimizer Description of subject: The Adam optimizer is a popular stochastic gradient descent method in machine learning that adaptively adjusts learning rates for each parameter using estimates of first and second moments of gradients.
Referenced by (9)
Full triples — surface form annotated when it differs from this entity's canonical label.
this entity surface form:
AdamW
this entity surface form:
AdamW optimizer
this entity surface form:
Adam optimization algorithm
this entity surface form:
Adam optimization algorithm