Ephemerent AtlasEphemerent Education
Foundations field atlasMathematics · Computing · PhysicsOpen · account optional

Ephemerent Atlas

Learn the world by tracing how its ideas connect.

A free, source-backed route through STEM: clear explanations, worked examples, deterministic practice, and laboratories that make you predict before they move.

Start at first principlesQuantities, ratios, and scaleComplete explanation · practice · scale bench

Reading local evidenceQuantities, ratios, and scaleThis is the first unseen objective in the current reviewed field-trial route.

  1. unseenQuantities, ratios, and scale
  2. unseenProcedures and state
  3. unseenMeasurement and uncertainty

Choose a route

Three entry points. One connected map.

Begin with the field that calls to you, or follow a bridge when another discipline gives the idea a sharper language.

01

Mathematics

Build quantities into functions, vectors, change, and evidence—then carry those structures into code and physical models.

Entry lesson
Quantities, ratios, and scale
Route
5 connected concepts
Open routeStart lesson
02

Computing

Make procedures explicit, reason about data and cost, then cross the boundary from software systems into evaluated machine learning.

Entry lesson
Procedures and state
Route
5 connected concepts
Open routeStart lesson
03

Physics

Begin with what can be measured, then explain motion through interaction, conservation, waves, and field models.

Entry lesson
Measurement and uncertainty
Route
5 connected concepts
Open routeStart lesson

The route is inspectable

Lines carry obligations.

Solid lines mean prerequisite dependence. Dotted curves are the release’s four actual cross-discipline bridges. Select any concept to inspect its record.

Atlas curriculum routesThree prerequisite routes in mathematics, computing, and physics, with cross-discipline bridges.bridgebridgebridgebridge01 / MathematicsQuantities, ratios, and scale01 · unseenFunctions as dependence02 · unseenVectors and coordinate change03 · unseenChange and accumulation04 · unseenProbability as evidence05 · unseen02 / ComputingProcedures and state01 · unseenData and invariants02 · unseenAlgorithms and cost03 · unseenOwnership and systems04 · unseenLearning systems and evaluation05 · unseen03 / PhysicsMeasurement and uncertainty01 · unseenMotion and interaction02 · unseenEnergy and conservation03 · unseenWaves and information04 · unseenFields, models, and limits05 · unseen
  1. Mathematics from quantity to uncertainty
    1. Quantities, ratios, and scale
    2. Functions as dependence
    3. Vectors and coordinate change
    4. Change and accumulation
    5. Probability as evidence
  2. Computing from procedure to learning systems
    1. Procedures and state
    2. Data and invariants
    3. Algorithms and cost
    4. Ownership and systems
    5. Learning systems and evaluation
  3. Physics from measurement to fields
    1. Measurement and uncertainty
    2. Motion and interaction
    3. Energy and conservation
    4. Waves and information
    5. Fields, models, and limits
Inspect the 16 map connections
  1. PrerequisiteQuantities, ratios, and scaleFunctions as dependencesource:atlas-method · editorial-draft:subject-review-pending
  2. PrerequisiteFunctions as dependenceVectors and coordinate changesource:atlas-method · editorial-draft:subject-review-pending
  3. PrerequisiteVectors and coordinate changeChange and accumulationsource:atlas-method · editorial-draft:subject-review-pending
  4. PrerequisiteChange and accumulationProbability as evidencesource:atlas-method · editorial-draft:subject-review-pending
  5. PrerequisiteProcedures and stateData and invariantssource:atlas-method · editorial-draft:subject-review-pending
  6. PrerequisiteData and invariantsAlgorithms and costsource:atlas-method · editorial-draft:subject-review-pending
  7. PrerequisiteAlgorithms and costOwnership and systemssource:atlas-method · editorial-draft:subject-review-pending
  8. PrerequisiteOwnership and systemsLearning systems and evaluationsource:atlas-method · editorial-draft:subject-review-pending
  9. PrerequisiteMeasurement and uncertaintyMotion and interactionsource:atlas-method · editorial-draft:subject-review-pending
  10. PrerequisiteMotion and interactionEnergy and conservationsource:atlas-method · editorial-draft:subject-review-pending
  11. PrerequisiteEnergy and conservationWaves and informationsource:atlas-method · editorial-draft:subject-review-pending
  12. PrerequisiteWaves and informationFields, models, and limitssource:atlas-method · editorial-draft:subject-review-pending
  13. BridgeFunctions as dependenceProcedures and statesource:atlas-method · editorial-draft:subject-review-pending
  14. BridgeVectors and coordinate changeMotion and interactionsource:atlas-method · editorial-draft:subject-review-pending
  15. BridgeProbability as evidenceLearning systems and evaluationsource:atlas-method · editorial-draft:subject-review-pending
  16. BridgeAlgorithms and costFields, models, and limitssource:atlas-method · editorial-draft:subject-review-pending
Inspect the full map

Vector Meridian / Field films

Long ideas, cut into inspectable routes.

Owned technical films join the graph as contextual evidence. Atlas loads no YouTube resource until you explicitly ask for the player.

43:25 · tutorialModern LLM's from scratch

A field-film introduction to the mathematical, statistical, and systems ideas used to build modern large language models. Catalog metadata is complete; owner-exported captions and independent claim review remain pending.

02How Kimi K3 Nearly Caught Claude Fable — Inside the 2.8T Model35:0603Who Really Invented AI? The History Everyone Gets Wrong92:0504Yann LeCun Thinks AI Is Learning the Wrong Way — JEPA Explained32:5405An AI Hacked Hugging Face to Cheat — Then Claude Mythos Escaped27:19Browse all seven films

Your evidence, not a score

Unseen. Developing. Demonstrated. Needs refresh.

Atlas records attempts, hints, item forms, and transfer locally. Demonstration needs two distinct unhinted forms including transfer or project evidence. A later miss asks for refresh; it does not erase prior work.

Your learning record stays in this browser. It is append-only, works offline, and makes no account, network, or model request.

Canonical learning is complete without AI.No model call is required for lessons, labs, projects, or evidence.
Open releaseAI-use policyReview statusFree-user funded model usage: exactly $0