intrinsic-dimensionality-distinct-from-ambient

IN premise — summaries/2026/08/24/wiki-Nonlinear_dimensionality_reduction-chunk-1.md

Created 2026-08-24T17:11:20+00:00

Intrinsic dimensionality (the true number of independent generative parameters) can be far smaller than ambient dimensionality; e.g., a 1024-dimensional letter image manifold can have an intrinsic dimensionality of only 2 (rotation and scale).

Summary

The real complexity hiding behind a dataset is often a tiny fraction of the number of dimensions you see on the surface. A 1024-pixel letter image, for instance, is really just two knobs — rotation and scale — doing all the work, which means systems can compress, search, and reason in a much smaller space without losing anything essential.