Neural Tangent Kernel (NTK) & Infinite-Width Limits

A limit where wide neural networks behave like kernel methods: the model linearizes around initialization and training becomes kernel regression.

published · difficulty 4/5 · 18 min read

Reading map and next steps

Intuition

Build the mental picture first so the rest of the page has something to attach to.

PredictName the object in plain language, then predict what should change.Leave with one reusable mental picture before notation appears.

Neural networks can learn features: the representation changes during training.

The NTK perspective asks a different question:

What if the network is so wide that, during training, it barely moves in parameter space?

If parameters move only a little, the network behaves like its own first-order Taylor expansion around initialization. In that regime, training can often be analyzed as linearized dynamics rather than nonlinear feature learning. It becomes a kernel method in disguise, where the "features" are the gradients of the network output with respect to its parameters.

This is useful because kernel methods are mathematically tractable: you can often predict learning dynamics, and in some studied settings reason about generalization-related behavior, without simulating full training.

Section prompt

Copy-only prompts — each action copies a page-grounded prompt to your clipboard. Nothing is sent by this site.

Math

Translate the story into symbols, assumptions, and a derivation you can inspect.

InspectTrack the same object through the notation and check each symbol.Leave with the invariant the equations preserve.

Equation 1Inspect equation object\Theta(x,x') = \nabla_\theta f_\theta(x)^T \nabla_\theta f_\theta(x').Equation 2Inspect equation object\partial_t f_t(x) = -\sum_i \Theta(x,x_i) (f_t(x_i) - y_i),

Let $f_\theta(x)$ be a network and $\theta$ its parameters. The neural tangent kernel is:

\Theta(x,x') = \nabla_\theta f_\theta(x)^\top \nabla_\theta f_\theta(x').

Consider gradient flow on a squared loss over training points $(x_i,y_i)$ , with $\mathcal L(\theta) = \tfrac12\sum_i (f_\theta(x_i) - y_i)^2$ .

In the infinite-width NTK limit for the standard parameterizations studied in this literature, $\Theta$ becomes deterministic and stays approximately constant during training. The function evolves according to:

\partial_t f_t(x) = -\sum_i \Theta(x,x_i)\,(f_t(x_i) - y_i),

which is a linear ODE in function space. The solution corresponds to kernel regression with kernel $\Theta$ (with details depending on loss, step size, and regularization).

Section prompt

Copy-only prompts — each action copies a page-grounded prompt to your clipboard. Nothing is sent by this site.

Code

Keep the implementation aligned with the notation so the algorithm is legible.

TraceMatch variables to symbols before reading the implementation.Leave with a runnable witness for the math.

Code witness 1import numpy as np rs = np.random.RandomState(0) m = 128 # "width" w = rs.randn(m) v = rs.ran...python

import numpy as np

rs = np.random.RandomState(0)
m = 128  # "width"
w = rs.randn(m)
v = rs.randn(m)

def grad_f(x):
    z = w * x
    s = (z > 0).astype(float)
    g_v = np.maximum(z, 0) / np.sqrt(m)
    g_w = (v * s * x) / np.sqrt(m)
    return np.concatenate([g_v, g_w])

def K(x, xp):
    g1, g2 = grad_f(x), grad_f(xp)
    return float(g1 @ g2)

xtr = np.linspace(-1, 1, 40)
ytr = np.sin(3 * xtr)
Ktr = np.array([[K(x, xp) for xp in xtr] for x in xtr])

lam = 1e-3
alpha = np.linalg.solve(Ktr + lam * np.eye(len(xtr)), ytr)

xte = np.linspace(-1, 1, 200)
Kte = np.array([[K(x, xp) for xp in xtr] for x in xte])
pred = Kte @ alpha
mse = np.mean((pred - np.sin(3 * xte)) ** 2)
print("test MSE:", round(float(mse), 4))

Section prompt

Copy-only prompts — each action copies a page-grounded prompt to your clipboard. Nothing is sent by this site.

Interactive Demo

Use direct manipulation to connect the explanation to a moving system.

ManipulateChange one control and predict the visible response before reveal.Leave with the observed invariant or a repaired model.

Live Concept Demo

Explore Neural Tangent Kernel (NTK) & Infinite-Width Limits

The stage is code-native and interactive. Use it to test the explanation against the mechanism.

difficulty 4/5graduatecode-aligned

Demo inquiry checkpoint

Manipulate one control and predict the visible change.

01Choose lensTrace a quantity

02ObserveDemo state pending

03GroundName the equation, invariant, or control that explains it.

04CarryNext: Scaling Laws & Emergent Abilities

Choose what to inspect in Neural Tangent Kernel (NTK) & Infinite-Width Limits. This shared fallback is an observation guide, not evidence of learning.

Loading interactive demo...

Use the demo to compare:

kernel-like learning dynamics (fixed features / linearized training),
feature learning (representations change), and
how width and fixed-feature assumptions push you toward one regime or the other.

Section prompt

Copy-only prompts — each action copies a page-grounded prompt to your clipboard. Nothing is sent by this site.

Object

4/4 sections ready

Concept: Neural Tangent Kernel (NTK) & Infinite-Width Limits

What is the smallest example that makes Neural Tangent Kernel (NTK) & Infinite-Width Limits click without losing the math?

BeforeOverparameterization & Generalization (Double Descent)Now4/4 sections readyTryManipulate one control and predict the visible change.NextScaling Laws & Emergent Abilities

Object contextScaling

StateScaling

2 sources attachedChecking local snapshot

concept:scaling/ntk

Math2 equation objects\Theta(x,x') = \nabla_\theta f_\theta(x)^T \nabla_\theta f_\theta(x').CodeNeural Tangent Kernel (NTK) & Infinite-Width Limits code witness 1rs = np.random.RandomState(0)DemoNeural Tangent Kernel (NTK) & Infinite-Width Limits predictionManipulate one control and predict the visible change.

ConceptLearner lens

Neural Tangent Kernel (NTK) & Infinite-Width Limits

What is the smallest example that makes Neural Tangent Kernel (NTK) & Infinite-Width Limits click without losing the math?

Mode questionCan I say the mechanism back in one sentence before I reveal anything?

Start with the prediction checkpoint, then compare the reveal to the mental model.

Take this move

Study modes

Keep the object fixed; change the lens.

01PredictCommit before reveal.02ExplainName the invariant.03Ground2 source links nearby.04ExperimentChange one knob.05CarrySave the next question.

Route back through the notebook

Carry the same object through intuition, math, code, and demo.

4/4 sections ready

Carry inOverparameterization & Generalization (Double Descent)

Bring the mental model from Overparameterization & Generalization (Double Descent); this page will reuse it instead of restarting from zero.

Work hereNeural Tangent Kernel (NTK) & Infinite-Width Limits

A limit where wide neural networks behave like kernel methods: the model linearizes around initialization and training becomes kernel regression.

Carry outScaling Laws & Emergent Abilities

The next edge should feel earned: use the demo prediction here before following Scaling Laws & Emergent Abilities.

After The First Pass

Turn the concept into an inspected object.

The lower panels are one second act: keep the object fixed, inspect it visually, check source boundaries, practice transfer, then attach the research question.

ConceptNeural Tangent Kernel (NTK) & Infinite-Width LimitsScaling

01MechanismInspect motionSee how the idea changes form before adding more explanation.02VisualChoose a cueBind one visible mark to one mathematical role.03SourcesCheck boundary2 source links attached to the object.04ClaimAudit evidenceTreat the page claim as something to verify.05PracticeAttempt transferWrite before feedback so the model has something to repair.06RoomAsk groundedAttach the question to the same exact object.

Mechanism Storyboard

See the idea move before the page explains it

A limit where wide neural networks behave like kernel methods: the model linearizes around initialization and training becomes kernel regression.

Demo notes open01 / Intuition

Editorial scaling illustration of a neural tangent kernel matrix, infinite-width curve, and function-space dynamics.

Prediction lens

Start with the picture, metaphor, or geometric mechanism.

Commit first

Before reading further, choose the kind of change Neural Tangent Kernel (NTK) & Infinite-Width Limits should make visible.

Visual Inquiry

Make the image answer a mathematical question

A limit where wide neural networks behave like kernel methods: the model linearizes around initialization and training becomes kernel regression.

4/4 stages readyDemo notes connected

Visual cueWhich visible object should carry the first intuition?

Inspection depth2/4

Prediction

Which visible object should carry the first intuition?

Commit first

Pick the cue that should make Neural Tangent Kernel (NTK) & Infinite-Width Limits easier to reason about before the page gives the answer.

Source Grounding

Canonical references for the mechanism on this page.

Object - ConceptNeural Tangent Kernel (NTK) & Infinite-Width LimitsQuestion

What is the smallest example that makes Neural Tangent Kernel (NTK) & Infinite-Width Limits click without losing the math?

concept:scaling/ntk

Boundary

sources: jacot-2018-ntk, lee-2019-wide-networks-linear

Check

Open the closest source note before trusting the local explanation.

Evidence

2 selected-object sources shown first; 2 references total.

Next move

Audit the claim boundary, then ask from the same selected object.

selected object source · paper · 2018Neural Tangent Kernel: Convergence and Generalization in Neural NetworksJacot, Gabriel, and Hongler

Located CF editorial boundary

Introduces the neural tangent kernel as the infinite-width kernel governing gradient-descent dynamics.

Used here as

Jacot et al. introduce the NTK as the parameter-gradient inner-product kernel governing infinite-width gradient-flow dynamics. Lee et al. support the near-initialization linearized-traini...

Caveat

This does not certify finite-width feature learning, all architectures or parameterizations, all losses, all optimizers, or generalization guarantees. The checked claim is scoped to infin...

Open source

selected object source · paper · 2019Wide Neural Networks of Any Depth Evolve as Linear Models Under Gradient DescentLee et al.

Located CF editorial boundary

Grounds the linearized-training view of wide neural networks near initialization.

Used here as

Jacot et al. introduce the NTK as the parameter-gradient inner-product kernel governing infinite-width gradient-flow dynamics. Lee et al. support the near-initialization linearized-traini...

Caveat

This does not certify finite-width feature learning, all architectures or parameterizations, all losses, all optimizers, or generalization guarantees. The checked claim is scoped to infin...

Open source

Claim Review

A limit where wide neural networks behave like kernel methods: the model linearizes around initialization and training becomes kernel regression.

Object - ConceptNeural Tangent Kernel (NTK) & Infinite-Width LimitsQuestion

What is the smallest example that makes Neural Tangent Kernel (NTK) & Infinite-Width Limits click without losing the math?

concept:scaling/ntk

Boundary

sources: jacot-2018-ntk, lee-2019-wide-networks-linear

Check

Treat every claim as provisional until source support and a local witness agree.

Evidence

1 structured claim check on this concept.

Next move

Run the prediction or practice transfer before asking for a grounded review.

1 CF editorial source-scope review recorded

Publisher-side editorial review is not independent replication. Claims without it still need exact source-support review. 2 references and 3 local witnesses are available for inspection.

In infinite-width NTK limits, or when wide-network training is well approximated by linearization near initialization, the parameter-gradient inner-product NTK governs squared-loss gradient-descent function dynamics like a kernel method.

Registered CF editorial boundary

Neural Tangent Kernel: Convergence and Generalization in Neural NetworksIntroduces the neural tangent kernel as the infinite-width kernel governing gradient-descent dynamics.Wide Neural Networks of Any Depth Evolve as Linear Models Under Gradient DescentGrounds the linearized-training view of wide neural networks near initialization.

Used here as

Jacot et al. introduce the NTK as the parameter-gradient inner-product kernel governing infinite-width gradient-flow dynamics. Lee et al. support the near-initialization linearized-training view for sufficie...

Local witness

source-span-jacot-2018-ntk source-span-lee-2019-wide-networks-linear math-object-1 math-object-2 code-witness-1 interactive-demo

Equation 1

\Theta(x,x') = \nabla_\theta f_\theta(x)^\top \nabla_\theta f_\theta(x').

Equation 2

\partial_t f_t(x) = -\sum_i \Theta(x,x_i)\,(f_t(x_i) - y_i),

Code witness 1import numpy as np rs = np.random.RandomState(0) m = 128 # "width" w = rs.randn(m) v = rs.ran...

Caveat

This does not certify finite-width feature learning, all architectures or parameterizations, all losses, all optimizers, or generalization guarantees. The checked claim is scoped to infinite-width or explici...

Review stateCF editorial source-scope reviewClaim metadata: source checkedPublisher-side editorial review only; not independent replication. Check caveats and exact source scope.

Reviewed source TeX: Jacot et al. support the NTK as the parameter-gradient kernel whose infinite-width limit becomes deterministic/constant and governs least-squares gradient-flow dynamics as kernel gradient descent; Lee et al. support the first-order Taylor/linearized near-initialization view for sufficiently wide networks. Local math/code/demo instantiate the fixed-gradient kernel ODE only in this scoped regime.

Reviewer: codex; reviewed 2026-05-20

Practice · Neural Tangent Kernel (NTK) & Infinite-Width Limits

Try the idea in your own words

A limit where wide neural networks behave like kernel methods: the model linearizes around initialization and training becomes kernel regression.

Concept · Current object

Neural Tangent Kernel (NTK) & Infinite-Width Limits

Source boundary: sources: jacot-2018-ntk, lee-2019-wide-networks-linear

Object context and links

Scaling

concept:scaling/ntk

Choose a task

Explain the mechanism

For Neural Tangent Kernel (NTK) & Infinite-Width Limits: What is the smallest example that makes Neural Tangent Kernel (NTK) & Infinite-Width Limits click without losing the math? Explain your answer, including what changes, why, and which assumption matters.

Your answerNo answer yet

A rough first thought is enough. Your draft stays when you change tasks.

Local to this page session. Not saved after leaving or reloading.

A little help · Explain

Open one hint at a time. These are suggestions, not your answer or a grade.

0 of 3 hints shown for this question.

Clearing your answer does not erase help history. Outside help cannot be verified here.

Where am I stuck? (optional)

Your own description, not an automatic diagnosis

Choose one, or leave this unspecified. Select it again to clear it.

Take your draft to a feedback conversation

No AI feedback runs here. You can copy a prompt to use elsewhere; nothing is sent automatically. Review the text before sharing, and leave out private information.

Write an attempt before copying a feedback prompt.

This draft and any AI response do not establish mastery. Try a different case later without help; this page has not measured that learning.

Grounded object roomClose

Selected object routeAsk from this object; carry one invariant back.sources: jacot-2018-ntk, lee-2019-wide-networks-linear

ObjectConceptNeural Tangent Kernel (NTK) & Infinite-Width Limits
PredictBefore revealNeural Tangent Kernel (NTK) & Infinite-Width Limits prediction
WitnessCompare codeNeural Tangent Kernel (NTK) & Infinite-Width Limits code witness 1
RoomAsk groundedChecking local snapshot

ConceptNeural Tangent Kernel (NTK) & Infinite-Width LimitsScaling

Code witness comparisonNeural Tangent Kernel (NTK) & Infinite-Width Limits code witness 1rs = np.random.RandomState(0)Prediction before revealNeural Tangent Kernel (NTK) & Infinite-Width Limits predictionManipulate one control and predict the visible change.

Grounded room questionWhat is the smallest example that makes Neural Tangent Kernel (NTK) & Infinite-Width Limits click without losing the math?Checking local snapshot

Research Room

Attach the question to an exact object

Pick the concept, equation, source, code witness, claim, misconception, or demo state before asking for help. The handoff stays grounded to that object.

Next local actionNo local draft saved yet

Open the draft below to save one note and next action in this browser.

conceptScaling

Neural Tangent Kernel (NTK) & Infinite-Width Limits

Anchored question

What is the smallest example that makes Neural Tangent Kernel (NTK) & Infinite-Width Limits click without losing the math?

Source boundaryInspect source ids: jacot-2018-ntk, lee-2019-wide-networks-linearStable content-object key attached

Role lenses for this object

These are fixed, deterministic perspectives derived from the selected object. They do not represent people, community contributions, or independent review.

Learner evidence requestAsk what would make "Neural Tangent Kernel (NTK) & Infinite-Width Limits" feel predictable rather than familiar.

Assumption

Source ids jacot-2018-ntk, lee-2019-wide-networks-linear must support the exact object, not just the surrounding topic.

Source-checking summary

Treat this as a mechanism object: connect the definition to one equation, code witness, or demo before broadening the discussion.

Proposed experiment

Ask the learner to perturb one representation, then check whether the same invariant survives in math, code, and demo.

Next action

The learner can state the mechanism in their own words

Evidence4 checks

PredictionChecking carried observation

ActionReady for one action

AILearner handoff ready

Open source object

01PredictionChecking browser-local route memory

02EvidenceChecking for a carried observation

03BoundaryInspect source ids: jacot-2018-ntk, lee-2019-wide-networks-linear

04Next moveSave one next action

Local action draftNo local draft saved yetExpand only when ready to capture one local next action

Local action draft

This draft stays locally in this browser for concept:scaling/ntk.

Draft noteNext action

No local draft saved.

Evidence to inspect

Source ids to inspect: jacot-2018-ntk, lee-2019-wide-networks-linear
Definition, prerequisite, and contrast concept links
The equation or code witness that makes the concept operational
One demo state that shows the invariant instead of a slogan

What would resolve this

The learner can state the mechanism in their own words
The learner can name the prerequisite that would repair confusion
The learner can predict how the mechanism changes under one perturbation

Object-attached AI handoff

I am working in Continuous Function's research reading room. Object: concept - Neural Tangent Kernel (NTK) & Infinite-Width Limits Object key: concept:scaling/ntk Context: Scaling Anchor id: concept/concept-notebook/scaling/ntk Open question: What is the smallest example that makes Neural Tangent Kernel (NTK) & Infinite-Width Limits click without losing the math? Evidence to inspect: - Source ids to inspect: jacot-2018-ntk, lee-2019-wide-networks-linear - Definition, prerequisite, and contrast concept links - The equation or code witness that makes the concept operational - One demo state that shows the invariant instead of a slogan Deterministic role lenses for this object: - Boundary: fixed perspectives, not people, community contributions, or independent review - Source-checking summary: Treat this as a mechanism object: connect the definition to one equation, code witness, or demo before broadening the discussion. - Proposed experiment: Ask the learner to perturb one representation, then check whether the same invariant survives in math, code, and demo. - Teach/transfer move: Turn the mechanism into one sentence that predicts a neighboring concept. - Assumptions: - Source ids jacot-2018-ntk, lee-2019-wide-networks-linear must support the exact object, not just the surrounding topic. - The stable content-object key lets local drafts, prompts, and route memory attach without changing the source page. - The concept explanation is local atlas prose until checked against its math, code, and source support. - Prerequisite gaps should become a repair route, not a reason to leave the object vague. - Role-lens requests: - Learner: ask for "Ask what would make "Neural Tangent Kernel (NTK) & Infinite-Width Limits" feel predictable rather than familiar." | assumption: Source ids jacot-2018-ntk, lee-2019-wide-networks-linear must support the exact object, not just the surrounding topic. | next action: The learner can state the mechanism in their own words - Researcher: ask for "Source ids to inspect: jacot-2018-ntk, lee-2019-wide-networks-linear" | assumption: The stable content-object key lets local drafts, prompts, and route memory attach without changing the source page. | next action: The learner can name the prerequisite that would repair confusion - Experimenter: ask for "Choose one variable or condition to perturb before asking for an explanation." | assumption: The concept explanation is local atlas prose until checked against its math, code, and source support. | next action: The learner can predict how the mechanism changes under one perturbation - Professor: ask for "Find the smallest transferable rule a learner could reuse without the AI." | assumption: Prerequisite gaps should become a repair route, not a reason to leave the object vague. | next action: Teach or transfer: Turn the mechanism into one sentence that predicts a neighboring concept. What would resolve this: - The learner can state the mechanism in their own words - The learner can name the prerequisite that would repair confusion - The learner can predict how the mechanism changes under one perturbation Answer as a careful research tutor: stay source-grounded, separate verified evidence from assumptions, name the relevant math objects, and end with one next action. Current deterministic role lens for this object: - Role lens: Learner - Evidence request: Ask what would make "Neural Tangent Kernel (NTK) & Infinite-Width Limits" feel predictable rather than familiar. - Assumption to keep visible: Source ids jacot-2018-ntk, lee-2019-wide-networks-linear must support the exact object, not just the surrounding topic. - Proposed experiment: Ask the learner to perturb one representation, then check whether the same invariant survives in math, code, and demo. - Next action: The learner can state the mechanism in their own words

concept/concept-notebook/scaling/ntk
concept:scaling/ntk

Return to reading

Reading map

In this concept

Manipulate one control and predict the visible change.

Try the demo

Continue with

Scaling Laws & Emergent Abilities

Connections and sources

Domain

Scaling

scaling
theory
ntk
kernels

Prerequisites

Overparameterization & Generalization (Double Descent)

Leads To

Scaling Laws & Emergent Abilities

Object Companion

Ask beside the selected object

A limit where wide neural networks behave like kernel methods: the model linearizes around initialization and training becomes kernel regression.

Your question

GoalComfortStyleStuck on

Context prompt

You are my AI learning companion for Continuous Function. Current context: Scaling concept. Learning surface: Neural Tangent Kernel (NTK) & Infinite-Width Limits. What this page says: A limit where wide neural networks behave like kernel methods: the model linearizes around initialization and training becomes kernel regression. Current section: Intuition, math, code, and interactive demo. Suggested next step: Manipulate one control and predict the visible change.. Learner goal: Understand the idea. Learner comfort level: New to this. Preferred explanation style: Visual first. Task: Explain the central idea in plain language, then restate it with the exact math objects from the page. Answer in a way that helps me learn: ask one clarifying question only if needed, use intuition before notation, and end with one thing I should try on the page.

Copy-only tool: this panel composes a prompt for an AI tool you already use. Nothing you enter here is sent from this page.