Pith Number

pith:RA2L3U7S

pith:2026:RA2L3U7SF5FSQ4DDFLXTXW6AQG

not attested not anchored not stored refs resolved

How Data Augmentation Shapes Neural Representations

Alex H. Williams, Sarah E. Harvey, Tianxiao He

Data augmentation steers neural representations along distinct, predictable trajectories in an invariant shape space.

arxiv:2605.15306 v1 · 2026-05-14 · cs.LG · stat.ML

Open paper page JSON Open Graph Bundle Merged state Verified badge What is a Pith Number?

Add to your LaTeX paper

\usepackage{pith}
\pithnumber{RA2L3U7SF5FSQ4DDFLXTXW6AQG}

Prints a linked badge after your title and injects PDF metadata. Compiles on arXiv. Learn more · Embed verified badge

Record completeness

1 Bitcoin timestamp

2 Internet Archive

3 Author claim open · sign in to claim

4 Citations open

5 Replications open

✓ Portable graph bundle live · download bundle · merged state

The bundle contains the canonical record plus signed events. A mirror can host it anywhere and recompute the same current state with the deterministic merge algorithm.

Claims

C1strongest claim

Increasing augmentation strength leads to well-behaved trajectories in this space, and that different augmentation types steer representations in distinct directions. Moreover, insights from neural geometry can predict which representations provide the most improvement when ensembling models.

C2weakest assumption

That the chosen invariant metric embedding of hidden representations captures the geometric properties most relevant to generalization and ensembling performance.

C3one line summary

Data augmentation produces well-behaved trajectories in shape-invariant representation space, with augmentation type steering distinct directions and geometry predicting ensembling gains.

References

29 extracted · 29 resolved · 6 Pith anchors

[1] month = apr, year = 2018 · arXiv:1803.09820

[2] Khoshgoftaar, A survey on Image Data Augmentation for Deep Learning, J Big Data, 6 (2019), pp 2019 · doi:10.1186/s40537-019-0197-0

[3] A Kernel Theory of Modern Data Augmentation 2019 · arXiv:1803.06084

[4] Understanding Image Representations by Measuring Their Equivariance and Equivalence 2015

[5] A group-theoretic framework for data augmentation 2020

Formal links

2 machine-checked theorem links

Receipt and verification

First computed	2026-05-20T00:00:51.755550Z
Builder	pith-number-builder-2026-05-17-v1
Signature	Pith Ed25519 (`pith-v1-2026-05`) · public key
Schema	pith-number/v1.0

Canonical hash

8834bdd3f22f4b2870632aef3bdbc081abecf49f8714f57920a27b38a7e95aa3

Aliases

arxiv: 2605.15306 · arxiv_version: 2605.15306v1 · doi: 10.48550/arxiv.2605.15306 · pith_short_12: RA2L3U7SF5FS · pith_short_16: RA2L3U7SF5FSQ4DD · pith_short_8: RA2L3U7S

Agent API

Resolver JSON Graph JSON Events JSON Schema Signing key

Verify this Pith Number yourself

curl -sH 'Accept: application/ld+json' https://pith.science/pith/RA2L3U7SF5FSQ4DDFLXTXW6AQG \
  | jq -c '.canonical_record' \
  | python3 -c "import sys,json,hashlib; b=json.dumps(json.loads(sys.stdin.read()), sort_keys=True, separators=(',',':'), ensure_ascii=False).encode(); print(hashlib.sha256(b).hexdigest())"
# expect: 8834bdd3f22f4b2870632aef3bdbc081abecf49f8714f57920a27b38a7e95aa3

Canonical record JSON

{
  "metadata": {
    "abstract_canon_sha256": "6fb228d36b303dc0fc7e159fa17f9bfe859ca9a15431dea37dcd728eb523e3a2",
    "cross_cats_sorted": [
      "stat.ML"
    ],
    "license": "http://creativecommons.org/licenses/by/4.0/",
    "primary_cat": "cs.LG",
    "submitted_at": "2026-05-14T18:19:40Z",
    "title_canon_sha256": "3f60ba417194fd8793176695cc9493b0b65619759fd7e0d4a7a5c08afa5e4ff9"
  },
  "schema_version": "1.0",
  "source": {
    "id": "2605.15306",
    "kind": "arxiv",
    "version": 1
  }
}