Pith Number

pith:BFPBRD4F

pith:2022:BFPBRD4FKNT44DNQQDW5NDPNQJ

not attested not anchored not stored refs resolved

Teaching Models to Express Their Uncertainty in Words

Jacob Hilton, Owain Evans, Stephanie Lin

GPT-3 can learn to state its own uncertainty in natural language, and those statements map to well-calibrated probabilities.

arxiv:2205.14334 v2 · 2022-05-28 · cs.CL · cs.AI · 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{BFPBRD4FKNT44DNQQDW5NDPNQJ}

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 GPT-3 model can learn to express uncertainty about its own answers in natural language -- without use of model logits. These levels map to probabilities that are well calibrated.

C2weakest assumption

That the verbalized confidence levels reflect the model's actual epistemic uncertainty rather than surface-level imitation of training examples or prompt patterns.

C3one line summary

GPT-3 can learn to express well-calibrated uncertainty about its answers using natural language phrases rather than logits.

References

24 extracted · 24 resolved · 7 Pith anchors

[1] A General Language Assistant as a Laboratory for Alignment · arXiv:2112.00861

[2] https://www.gwern.net/GPT-3-nonfiction# calibration, Last accessed on 2022-04-24. Tom B. Brown, Benjamin Mann, Nick Ryder, Melanie Subbiah, Jared Kaplan, Prafulla Dhariwal, Arvind Neelakantan, Pranav 2022

[3] PaLM: Scaling Language Modeling with Pathways · arXiv:2204.02311

[4] Gabriela Csurka 2022

[5] arXiv preprint arXiv:1702.05374 (2017) https://doi.org/10.1007/ 978-3-319-58347-1 1 2020 · arXiv:1702.05374

Formal links

1 machine-checked theorem link

Cited by

29 papers in Pith

Towards Expert-Level Medical Question Answering with Large Language Models

Vision-Language and Large Language Model Performance in Gastroenterology: GPT, Claude, Llama, Phi, Mistral, Gemma, and Quantized Models

Trust or Abstain? A Self-Aware RAG Approach

Model-Adaptive Tool Necessity Reveals the Knowing-Doing Gap in LLM Tool Use

Position: Uncertainty Quantification in LLMs is Just Unsupervised Clustering

Receipt and verification

First computed	2026-05-17T23:38:47.109011Z
Builder	pith-number-builder-2026-05-17-v1
Signature	Pith Ed25519 (`pith-v1-2026-05`) · public key
Schema	pith-number/v1.0

Canonical hash

095e188f855367ce0db080edd68ded82449f590a6aed7fe6047344d1874fd1e7

Aliases

arxiv: 2205.14334 · arxiv_version: 2205.14334v2 · doi: 10.48550/arxiv.2205.14334 · pith_short_12: BFPBRD4FKNT4 · pith_short_16: BFPBRD4FKNT44DNQ · pith_short_8: BFPBRD4F

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/BFPBRD4FKNT44DNQQDW5NDPNQJ \
  | 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: 095e188f855367ce0db080edd68ded82449f590a6aed7fe6047344d1874fd1e7

Canonical record JSON

{
  "metadata": {
    "abstract_canon_sha256": "77c7c1f6c18f4d94025f67e8ed0712faa7b963a1b015c5b72a76f2e1043da9fa",
    "cross_cats_sorted": [
      "cs.AI",
      "cs.LG"
    ],
    "license": "http://arxiv.org/licenses/nonexclusive-distrib/1.0/",
    "primary_cat": "cs.CL",
    "submitted_at": "2022-05-28T05:02:31Z",
    "title_canon_sha256": "c0ee3606ce5cf484b933e412df2861c63b35d8078aa38ec6fbdc9c3ed3ce0a0e"
  },
  "schema_version": "1.0",
  "source": {
    "id": "2205.14334",
    "kind": "arxiv",
    "version": 2
  }
}