Pith Number

pith:PS6VCTOF

pith:2026:PS6VCTOFRVZ3CHJTX2QS2OSGIF

not attested not anchored not stored refs pending

A Controlled Counterexample to Strong Proxy-Based Explanations of OOD Performance: in a Fixed Pretraining-and-Probing Setup

Hongmin Li

A proxy for total learned structure can fail to track the task-relevant structure that drives OOD performance, even in a controlled setting.

arxiv:2605.11554 v2 · 2026-05-12 · cs.LG

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

Add to your LaTeX paper

\usepackage{pith}
\pithnumber{PS6VCTOFRVZ3CHJTX2QS2OSGIF}

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

A proxy for total learned structure can fail to track the task-relevant structure that drives OOD performance, even in a controlled setting.

C2weakest assumption

That the synthetic sequence-model experiment and the controlled construction adequately capture the dynamics of real-world pretraining and OOD probing scenarios.

C3one line summary

Proxy rankings of pretraining datasets by learned structure can reverse the actual OOD accuracy rankings in a synthetic sequence modeling task.

Formal links

2 machine-checked theorem links

Receipt and verification

First computed	2026-07-01T01:17:51.651642Z
Builder	pith-number-builder-2026-05-17-v1
Signature	Pith Ed25519 (`pith-v1-2026-05`) · public key
Schema	pith-number/v1.0

Canonical hash

7cbd514dc58d73b11d33bea12d3a4641795a457739a328ddf450e7061ef243fa

Aliases

arxiv: 2605.11554 · arxiv_version: 2605.11554v2 · doi: 10.48550/arxiv.2605.11554 · pith_short_12: PS6VCTOFRVZ3 · pith_short_16: PS6VCTOFRVZ3CHJT · pith_short_8: PS6VCTOF

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/PS6VCTOFRVZ3CHJTX2QS2OSGIF \
  | 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: 7cbd514dc58d73b11d33bea12d3a4641795a457739a328ddf450e7061ef243fa

Canonical record JSON

{
  "metadata": {
    "abstract_canon_sha256": "1bdbf345b04c3c44cdde230af57f333cb5deea7056fa1f4cb1dea743ebe74e3d",
    "cross_cats_sorted": [],
    "license": "http://arxiv.org/licenses/nonexclusive-distrib/1.0/",
    "primary_cat": "cs.LG",
    "submitted_at": "2026-05-12T05:36:46Z",
    "title_canon_sha256": "2989677a8541b601a6dc65d300fcb1834af648d06bccbc9b5db963f9b9ab2115"
  },
  "schema_version": "1.0",
  "source": {
    "id": "2605.11554",
    "kind": "arxiv",
    "version": 2
  }
}