Pith Number

pith:ASTJHJTF

pith:2020:ASTJHJTFV6SPSM2W3YA3UAD2VF

not attested not anchored not stored refs resolved

How Much Knowledge Can You Pack Into the Parameters of a Language Model?

Adam Roberts, Colin Raffel, Noam Shazeer

Fine-tuned language models answer questions using only knowledge stored in their parameters.

arxiv:2002.08910 v4 · 2020-02-10 · cs.CL · 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{ASTJHJTFV6SPSM2W3YA3UAD2VF}

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

fine-tuning pre-trained models to answer questions without access to any external context or knowledge ... performs competitively with open-domain systems that explicitly retrieve answers from an external knowledge source

C2weakest assumption

That the knowledge required to answer the questions is present in the pre-training data and can be effectively stored and retrieved from the model parameters through fine-tuning on QA pairs.

C3one line summary

Fine-tuned language models store knowledge in parameters to answer questions competitively with retrieval-based open-domain QA systems.

References

60 extracted · 60 resolved · 18 Pith anchors

[3] A discrete hard

[4] Chen, Danqi and Fisch, Adam and Weston, Jason and Bordes, Antoine , journal=. Reading

[5] Learning to Retrieve Reasoning Paths over

[6] Devlin, Jacob and Chang, Ming-Wei and Lee, Kenton and Toutanova, Kristina , journal=

[7] Yang, Zhilin and Dai, Zihang and Yang, Yiming and Carbonell, Jaime and Salakhutdinov, Ruslan and Le, Quoc V. , journal=

Formal links

2 machine-checked theorem links

Cited by

23 papers in Pith

Atlas: Few-shot Learning with Retrieval Augmented Language Models

ARIA: Adaptive Retrieval Intelligence Assistant -- A Multimodal RAG Framework for Domain-Specific Engineering Education

Dense Passage Retrieval for Open-Domain Question Answering

Tug-of-War within A Decade: Conflict Resolution in Vulnerability Analysis via Teacher-Guided Retrieval-Augmented Generations

SmolLM2: When Smol Goes Big -- Data-Centric Training of a Small Language Model

Receipt and verification

First computed	2026-05-17T23:38:53.873003Z
Builder	pith-number-builder-2026-05-17-v1
Signature	Pith Ed25519 (`pith-v1-2026-05`) · public key
Schema	pith-number/v1.0

Canonical hash

04a693a665afa4f93356de01ba007aa954c12eac308743f5150934ff9a119141

Aliases

arxiv: 2002.08910 · arxiv_version: 2002.08910v4 · doi: 10.48550/arxiv.2002.08910 · pith_short_12: ASTJHJTFV6SP · pith_short_16: ASTJHJTFV6SPSM2W · pith_short_8: ASTJHJTF

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/ASTJHJTFV6SPSM2W3YA3UAD2VF \
  | 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: 04a693a665afa4f93356de01ba007aa954c12eac308743f5150934ff9a119141

Canonical record JSON

{
  "metadata": {
    "abstract_canon_sha256": "da783045f046730f4f7d55bb7366b33d9c4a68171a5e2b5cc93ce32da568aa65",
    "cross_cats_sorted": [
      "cs.LG",
      "stat.ML"
    ],
    "license": "http://arxiv.org/licenses/nonexclusive-distrib/1.0/",
    "primary_cat": "cs.CL",
    "submitted_at": "2020-02-10T18:55:58Z",
    "title_canon_sha256": "2963317b125562b73a72cb6679954121a20b97cabd0104f1364ebc3b3f33a2c1"
  },
  "schema_version": "1.0",
  "source": {
    "id": "2002.08910",
    "kind": "arxiv",
    "version": 4
  }
}