Learn

Learning preferences

Make learning more comfortable. Changes apply immediately.

Enlarges page content, including diagrams and their labels. You can also use your browser’s zoom.

Choose the colours that are easiest for you to distinguish. Applies to lesson diagrams and lab plots. Labels, shapes, hatching and line patterns also distinguish meanings. Source photographs and raw image brightness stay unchanged; mask-only views are available for image overlays.

ExampleComparisonSample link

Saved in this browser on this device. Preferences are separate from progress backups.

All courses

Find the idea behind the idea.

Explore a foundation, then return to your lesson. The full AI lesson map below connects all 16 modules. Open a topic’s title to learn it, or choose “Explore connections” to follow its links through the map. Every lesson is open to you.

CONNECTED IDEAS · SMALL PILOT CLUSTER

Explore one idea at a time

Use a connection to review, then return. These links never lock lessons or award mastery.

  1. Slope

    Gradient descent uses as a foundation Slope.

    To explain the direction of this scalar gradient update, interpret the local slope of loss with respect to weight. A negative slope suggests increasing the weight for a sufficiently small step.

    Graphs and coordinatesChain rule and gradientsExplore these connections
    Scope and source review

    Explaining the scalar loss update in the pilot, not a course enrollment requirement or a claim that any finite step improves loss.

    Source-audited 2026-10-31. Independent subject-expert and learner review remain pending.

  2. Learning rate

    Gradient descent can review Learning rate.

    The gradient supplies direction and scale; the learning rate multiplies it. Comparing proposed steps with actual losses explains overshoot.

    Use momentum and adaptive stepsFollow the slope of a mistakeExplore these connections
    Scope and source review

    One bounded scalar update and its actual loss; not a convergence guarantee.

    Source-audited 2026-10-31. Independent subject-expert and learner review remain pending.

  3. Neural networks

    Gradient descent applies to Neural networks.

    Once gradients are calculated, an optimizer can update network parameters. Backpropagation computes sensitivities; it is not itself the update.

    Generate by removing noiseUnderstand Neural NetworksExplore these connections
    Scope and source review

    The simultaneous two-weight update in the backpropagation lesson.

    Source-audited 2026-10-31. Independent subject-expert and learner review remain pending.

11 concepts in this cross-subject pilot, not a complete curriculum graph. The existing AI lesson map and progress records remain separate and available below or through AI discovery.

Your AI learning map

Suggestions use your last lesson and connections to topics with recorded completion or a passed check. These are suggestions, not prerequisites or a readiness assessment. All topics remain freely accessible.

Loading your suggestions…

168 of 168 topics match. Filters affect the topic directory; connections always include other modules.

Choose a topic from the directory (168)

How does a network learn?

What has the network learned?

How does AI process language?

Why does AI need computation?

Data audits and training diagnosis

Simple models and useful baselines

Generative models beyond text

Learning through actions

Planning and reasoning

Learning without losing old skills

Connecting tools and specialists

Memory and shared working state

Choosing what to do next

Testing whether improvement is real

Assemble an experimental agent

Geometry and control foundations

Connections for Your first neuron

Solid arrows show foundations → selected topic → lessons that build on it. Dashed arrows show suggested applications. Labels and line patterns distinguish the connections in every palette.

Read connections and explanations

On smaller screens, scroll the map horizontally, or use the text view below. Tab moves between topic links.

1 · Builds on
2 · Selected topic
3 · Enables / applications
Refresher / practice Algebra Open the linked practice
You are here · How does a network learn?Your first neuron
Loading progress…
Open lesson

How this topic connects

Use these connections to review a foundation or choose where to go next. You can open any topic without completing the others.

What it builds on

  • Algebra

    Substitute input values into a formula and keep track of which quantities change. This makes the lesson’s worked calculations easier to follow.

You are hereYour first neuron

Turn inputs, a weight, and a bias into a prediction.

What it enables