data-geometry-not-capacity-determines-architecture

IN derived (depth 3)

Created 2026-06-21T10:13:05+00:00 · Reviewed 2026-06-21T15:37:01+00:00

Architecture design is governed by data geometry rather than model capacity — the manifold hypothesis explains why inductive biases (local connectivity, weight sharing) succeed, while the depth bottleneck reveals that theoretical capacity (universal approximation) matters less than trainability (residual connections enabling gradient flow through the data manifold's structure).

Justifications

SL — depth-3 — both depth-2 conclusions independently point to the same principle: what matters is matching architecture to data structure, not maximizing representational power

Antecedents (all must be IN):

  • IN manifold-structure-explains-inductive-bias-effectiveness — The manifold hypothesis offers a theoretical lens for understanding why certain architectural inductive biases are effective — if high-dimensional data lies on low-dimensional manifolds (and prediction is formally equivalent to compression of that structure), then architectures exploiting local connectivity and weight sharing can be seen as responses to data geometry rather than purely ad hoc engineering choices.
  • IN trainability-not-capacity-was-the-depth-bottleneck — The theory-practice gap in neural network depth — where universal approximation says shallow networks suffice but practice demands deep ones — reveals that the real bottleneck was never representational capacity but trainability, and residual connections resolved exactly this optimization barrier, enabling the hierarchical representations that shallow networks could theoretically but not practically learn.

Dependents

These beliefs depend on this one: