---
title: "Birdal et al.'s persistent-homology-dimension estimator subsamples training iterates uniformly at random, not by any structural covariate"
type: "claim"
status: "seedling"
writer_model: "claude-sonnet-5"
source_url: "https://arxiv.org/abs/2111.13171"
source_title: "Intrinsic Dimension, Persistent Homology and Generalization in Neural Networks"
source_author: "Tolga Birdal, Aaron Lou, Leonidas Guibas, Umut Şimşekli"
source_date: "2021-11"
source_quote: "Wn ← sample(W, n) // random sampling"
source_tier: 1
audit_status: "capture-verified — full PDF fetched and read directly via extract_pdf at capture time (arxiv.org/pdf/2111.13171, tls:verified), not an abstract-only read; not independently re-fetched at this promotion pass, since the capture's method (direct full-text read of Algorithm 1 and the Section 5.2 ablation, not a secondary summary) is already stronger evidence than the abstract-only verification the sibling claim-note [[claim-birdal-2021-persistent-homology-dimension-bounds-generalization]] currently carries."
provenance: "Promotion from 10-inbox/raw/2026-07-21-do-persistent-homology-based-neural-network-generalization-diagnostics.md, 2026-07-22 (headless)"
origin: "batch"
derived_from: "10-inbox/raw/2026-07-21-do-persistent-homology-based-neural-network-generalization-diagnostics.md"
date_created: "2026-07-22T00:00:00.000Z"
tags: ["persistent-homology","topological-data-analysis","sampling-artifact","generalization","neural-networks","subsampling"]
audits: ["2026-07-23 claude-fable-5"]
drafted_in: ["what-the-gradient-cant-see"]
---


Birdal, Lou, Guibas & Şimşekli's PHD estimator
([[claim-birdal-2021-persistent-homology-dimension-bounds-generalization]]) is
computed by an algorithm whose own pseudocode names its sampling step
explicitly: "Wn ← sample(W, n) // random sampling." The experimental section
confirms the same procedure in prose — "For varying n, we randomly pick n of
last iterates and compute Eα0" — repeatedly drawing subsets of the recorded
training trajectory at random, never by a structural property such as weight
magnitude, layer, or training-loss rank.

The paper does document a known estimator bias tied to sample *size*: in its
synthetic ground-truth ablation, "as n increases our estimator becomes
smoother and well approximates the GT up to a slight over-estimation, a
repeatedly observed phenomenon" (citing the intrinsic-dimension-estimation
survey literature). This is a magnitude bias that shrinks monotonically as
sample size grows — a materially different failure mode from a
sign-reversing artifact tied to *which* points are chosen, of the kind
[[claim-hub-selection-artifact-can-reverse-network-breakpoint-signal]]
documents in a degree-heterogeneous trade network.

Because PHD's only sampling axis is uniform-random-over-time, the hub-selection
mechanism has no structural-covariate entry point to exploit here — see
[[observation-hub-selection-artifact-absent-by-design-in-founding-ph-generalization-papers]]
for the fuller synthesis across both founding NN-PH diagnostics, which answers
[[question-tda-neural-net-sampling-artifact-risk]] for the two papers it named.