Pith Number

pith:KOV64XHE

pith:2023:KOV64XHEEHE6X7VXUTHPFTDZDN

not attested not anchored not stored refs resolved

The Internal State of an LLM Knows When It's Lying

Amos Azaria, Tom Mitchell

The hidden activations inside an LLM can be read by a trained classifier to detect whether a statement is true or false.

arxiv:2304.13734 v2 · 2023-04-26 · 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{KOV64XHEEHE6X7VXUTHPFTDZDN}

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

the LLM's internal state can be used to reveal the truthfulness of statements. This includes both statements provided to the LLM, and statements that the LLM itself generates.

C2weakest assumption

That the hidden activations contain a generalizable signal of truthfulness that is not merely an artifact of the particular training sentences or superficial statistical properties shared with the labels.

C3one line summary

Hidden activations in LLMs encode detectable information about statement truthfulness, enabling a classifier to identify true versus false content more reliably than the model's assigned probabilities.

References

28 extracted · 28 resolved · 7 Pith anchors

[1] Llama 2: Early Adopters' Utilization of Meta's New Open-Source Pretrained Model , author=. 2023 , publisher= 2023

[5] Advances in neural information processing systems , volume=

[8] ACM Computing Surveys , volume= 2023

[11] Proceedings of the national academy of sciences , volume= 2017

[12] Handbook of Research on Machine Learning Applications and Trends: Algorithms, Methods, and Techniques , pages= 2010

Formal links

2 machine-checked theorem links

Cited by

27 papers in Pith

HalluScan: A Systematic Benchmark for Detecting and Mitigating Hallucinations in Instruction-Following LLMs

Exposing the Ghost in the Transformer: Abnormal Detection for Large Language Models via Hidden State Forensics

A Multi-Source Framework for Relational Validation of Large Language Models Using Expert-Curated Encyclopedic Sources

Mechanistic Interpretability Needs Philosophy

How do LLMs Compute Verbal Confidence

Receipt and verification

First computed	2026-05-17T23:38:49.657733Z
Builder	pith-number-builder-2026-05-17-v1
Signature	Pith Ed25519 (`pith-v1-2026-05`) · public key
Schema	pith-number/v1.0

Canonical hash

53abee5ce421c9ebfeb7a4cef2cc791b5c6a8f4035a3370225bb51b68e8f82ff

Aliases

arxiv: 2304.13734 · arxiv_version: 2304.13734v2 · doi: 10.48550/arxiv.2304.13734 · pith_short_12: KOV64XHEEHE6 · pith_short_16: KOV64XHEEHE6X7VX · pith_short_8: KOV64XHE

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/KOV64XHEEHE6X7VXUTHPFTDZDN \
  | 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: 53abee5ce421c9ebfeb7a4cef2cc791b5c6a8f4035a3370225bb51b68e8f82ff

Canonical record JSON

{
  "metadata": {
    "abstract_canon_sha256": "5ccf1164771c2d94ac9d1d74ea147cf9ffb616d551ce05fe36a5b7fc0e739517",
    "cross_cats_sorted": [
      "cs.AI",
      "cs.LG"
    ],
    "license": "http://creativecommons.org/licenses/by-nc-nd/4.0/",
    "primary_cat": "cs.CL",
    "submitted_at": "2023-04-26T02:49:38Z",
    "title_canon_sha256": "954ff13209d1dc258fabe7058532d9a1bc37bbaf282cae3137636fd652f41ee3"
  },
  "schema_version": "1.0",
  "source": {
    "id": "2304.13734",
    "kind": "arxiv",
    "version": 2
  }
}