The Trilogy

Manim animation — Structured Knowledge Accumulation, in three parts: I, layer-wise entropy reduction; II, entropic least action (η ≡ Δt); III, geodesic learning in Riemannian neural fields.