Pith Number

pith:T4EZEB7I

pith:2026:T4EZEB7ISHBNPGY6QBDFAQGC34

not attested not anchored not stored refs resolved

From Baselines to Transport Geodesics: Axiomatic Attribution via Optimal Generative Flows

Cenwei Zhang, Lei You, Lin Zhu, Manxi Lin

Aumann-Shapley line integrals along transport geodesics give unique and stable attributions.

arxiv:2603.05093 v2 · 2026-03-05 · cs.LG · cs.AI · cs.CV

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

Add to your LaTeX paper

\usepackage{pith}
\pithnumber{T4EZEB7ISHBNPGY6QBDFAQGC34}

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

For a fixed path, the Aumann-Shapley line integral is the unique attribution rule under standard fixed-path axioms and explicit coordinate-trace regularity. For path selection, minimizing kinetic action over flows that transport a reference distribution to the data distribution yields a transport-geodesic attribution principle.

C2weakest assumption

That the data-generating process can be adequately modeled by flows whose kinetic action minimization produces attribution paths that are meaningfully better than hand-designed interpolations.

C3one line summary

Transport-geodesic attribution via optimal generative flows selects principled paths for feature attributions by minimizing kinetic action.

References

11 extracted · 11 resolved · 6 Pith anchors

[1] Shapley explainability on the data manifold 2006

[3] Progressive Growing of GANs for Improved Quality, Stability, and Variation · arXiv:1710.10196

[4] Flow Matching for Generative Modeling · arXiv:2210.02747

[5] Flow Straight and Fast: Learning to Generate and Transfer Data with Rectified Flow · arXiv:2209.03003

[6] URL https://proceedings.neurips 2017

Receipt and verification

First computed	2026-05-18T02:44:30.892543Z
Builder	pith-number-builder-2026-05-17-v1
Signature	Pith Ed25519 (`pith-v1-2026-05`) · public key
Schema	pith-number/v1.0

Canonical hash

9f099207e891c2d79b1e80465040c2df2e08e87d2c961473635b641dec38465e

Aliases

arxiv: 2603.05093 · arxiv_version: 2603.05093v2 · doi: 10.48550/arxiv.2603.05093 · pith_short_12: T4EZEB7ISHBN · pith_short_16: T4EZEB7ISHBNPGY6 · pith_short_8: T4EZEB7I

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/T4EZEB7ISHBNPGY6QBDFAQGC34 \
  | 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: 9f099207e891c2d79b1e80465040c2df2e08e87d2c961473635b641dec38465e

Canonical record JSON

{
  "metadata": {
    "abstract_canon_sha256": "3449834761bc1d5b34989c5d4f141988770cb27b9d15651c2b0b417472d6a39c",
    "cross_cats_sorted": [
      "cs.AI",
      "cs.CV"
    ],
    "license": "http://arxiv.org/licenses/nonexclusive-distrib/1.0/",
    "primary_cat": "cs.LG",
    "submitted_at": "2026-03-05T12:05:20Z",
    "title_canon_sha256": "794f2236ed1f020148ddf398cc6d0a32199c5d4761c3395b51057741365078cc"
  },
  "schema_version": "1.0",
  "source": {
    "id": "2603.05093",
    "kind": "arxiv",
    "version": 2
  }
}