Addressed
A concept's identity is one string. It's the URL, the content file's path, and the
graph node id, all at once — down to the section level, e.g.
mechanics/noether-theorem#statement.
A personal knowledge base of physics and math: 87 concepts, each its own addressable page, wired together by a public, typed knowledge graph — browsable as prose or as a map.
Domain nodes sized by concept count. Solid lines are the domain hierarchy; dashed
lines are real cross-domain links pulled from the atlas's own graph — e.g.
math/tensors → maxwell-equations, or
electroweak-unification → maxwell-equations.
Physics Atlas isn't a course, and it isn't a wiki clone of an existing textbook. It's a working map — physics and math concepts each written as their own page: prose, build-time-rendered math, and, where it earns its place, an interactive 3D visualization. Every page exists to be found and connected, not scrolled through once.
A concept's identity is one string. It's the URL, the content file's path, and the
graph node id, all at once — down to the section level, e.g.
mechanics/noether-theorem#statement.
Pages link and transclude each other, so a definition gets written once and reused, with attribution, wherever it's needed. A link to a concept that doesn't exist yet isn't an error — it becomes an entry on the wanted list.
Prerequisites, generalizations, derivations, and mentions mined from the text
combine into one typed graph — an interactive map, and machine-readable at
/api/graph.json, per concept, and as llms.txt.
One String Theory entry is original work, not a textbook summary:
strings/theory-of-amir — a conjecture that a black hole collapses one
spatial dimension across its horizon (D → D−1). It's filed as a normal node, on the
same footing as Maxwell's equations.
“A model can already explain the standard version of a topic — the one that's been written a thousand times. What's left worth doing by hand is the harder version: a structured, personally-maintained map precise enough to hold your own additions to a field, not just the received material.”— the project's working assumption