Rotate exact hypercubes through 8D, then push into sampled analytical previews up to 50D without losing the real combinatorics.
A lightweight projection is running for this session. The tesseract remains visible as a live 2D construction, with the same mathematical notes continuing below.
Every figure in this atlas is generated from first principles, not drawn by hand. A centred n-cube is the set of points whose coordinates are each ±1, so it has 16 vertices in the current dimension. Two vertices are joined by an edge exactly when their coordinate vectors differ in a single coordinate (Hamming distance one), giving 32 edges.
Face vector for the dimension selected above — updates live. Above 8D the screen uses a sampled preview, while the counts remain for the full n-cube.
| k | face type | count |
|---|
Vertices double each step; higher faces follow the binomial rule.
| dim | name | V | E | squares | cells |
|---|
Full derivations are in docs/mathematics.md.
A tesseract is a four-dimensional Euclidean spatial object. It is not the same thing as spacetime. The distinction matters, and this atlas keeps it explicit.
Special relativity treats spacetime as three spatial dimensions plus one time dimension (3+1). But time is not interchangeable with a spatial axis: in the interval it enters with the opposite sign, so a 4-cube of space is genuinely different from 4D spacetime. General relativity then describes spacetime as a curved four-dimensional manifold.
One historical idea adds a fifth spatial dimension curled into a tiny compact circle, unifying electromagnetism with gravity in a single geometric picture. It is mathematically elegant — but no experiment has established such a dimension exists.
Some frameworks require ten or eleven total dimensions for internal mathematical consistency, with the extra ones compactified at minute scales. These are theoretical frameworks, not confirmed observations.
Astronomy already lives in high-dimensional spaces: stellar spectra, exoplanet population parameters, Gaia astrometric measurements, cosmological likelihood surfaces, and machine-learning feature spaces. We never see them directly — we reason through projections, slices, correlations, and reduced representations. That is the same instinct this atlas trains.
The visualiser shows mathematical projections, not photographs or direct views of extra dimensions. Visual beauty is not evidence. A simulation can build intuition; it cannot replace experimental validation. More detail is in docs/physics-and-astronomy.md.
I work on astronomical instrumentation, where high-dimensional spaces are routine — spectra, parameter grids, likelihood surfaces — yet we only ever inspect them through projections, slices, and reduced representations. I built Hyperdimensional Atlas to apply that same habit to a clean test case: the n-cube. Every figure here is generated directly from the polytope's definition, so what you rotate and slice is the exact object, shown as a projection rather than a direct view.
The interface keeps three things deliberately separate: the mathematics, which is exact; established physics, which is narrow; and speculation, which is labelled as such. Everything runs in your browser from plain HTML, CSS, and JavaScript modules with Three.js — no server, no tracking. More detail is in the README and the references.