Pith Number

pith:N7BATLN5

pith:2025:N7BATLN5YJWMBDF5XOT3BUFQ7X

not attested not anchored not stored refs pending

How do language models learn facts? Dynamics, curricula and hallucinations

Andrew Lampinen, J\"org Bornschein, Nicolas Zucchet, Razvan Pascanu, Soham De, Stephanie Chan

arxiv:2503.21676 v2 · 2025-03-27 · cs.CL · 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{N7BATLN5YJWMBDF5XOT3BUFQ7X}

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.

Cited by

9 papers in Pith

Natural Ungrokking: Asymmetric Control of Which Rules Survive Pretraining

Structure Before Collapse: Transient semantic geometry in next-token prediction

The Future of Facts: Tracing the Factual Generation-Verification Gap

Deep sequence models tend to memorize geometrically; it is unclear why

Fine-Tuning Without Forgetting via Loss-Adaptive Learning Rates

Receipt and verification

First computed	2026-07-05T11:42:28.026183Z
Builder	pith-number-builder-2026-05-17-v1
Signature	Pith Ed25519 (`pith-v1-2026-05`) · public key
Schema	pith-number/v1.0

Canonical hash

6fc209adbdc26cc08cbdbba7b0d0b0fdd7df22204eaa244b9c6c097d48c04f69

Aliases

arxiv: 2503.21676 · arxiv_version: 2503.21676v2 · doi: 10.48550/arxiv.2503.21676 · pith_short_12: N7BATLN5YJWM · pith_short_16: N7BATLN5YJWMBDF5 · pith_short_8: N7BATLN5

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/N7BATLN5YJWMBDF5XOT3BUFQ7X \
  | 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: 6fc209adbdc26cc08cbdbba7b0d0b0fdd7df22204eaa244b9c6c097d48c04f69

Canonical record JSON

{
  "metadata": {
    "abstract_canon_sha256": "9d313aff0067a1316b97e7a08ae2c8149650759e2ba645d0d1df981f9af8f5ff",
    "cross_cats_sorted": [
      "cs.LG"
    ],
    "license": "http://creativecommons.org/licenses/by/4.0/",
    "primary_cat": "cs.CL",
    "submitted_at": "2025-03-27T16:43:45Z",
    "title_canon_sha256": "c4d6664ada388834dad9f85b90c76255a8a5de9191822c9cd57584dfb67f459d"
  },
  "schema_version": "1.0",
  "source": {
    "id": "2503.21676",
    "kind": "arxiv",
    "version": 2
  }
}