Revisions

Each change we have recorded between two versions: the old and the new text, whether the document's changelog explains it, how we know, and whether we have checked it. A revision is not in itself evidence of wrongdoing; most are corrections (Methodology §7).

  1. Wording changed · 11 Jun 2026 to 25 Jun 2026

    Alignment risk in executive summary

    lowvery low

    Explained — corrected to match §2.4

    How we know Changelog read 26 Sep 2026

    Checked Confirmed from changelog

Versions

Four versions are on record, oldest first. We hold a copy of one; the others are known only by their date.

  1. 9 Jun 2026

    First published version

    Date from
    our source registry
    Copy
    Known to exist; no copy held
  2. 11 Jun 2026

    Revision

    Date from
    the document's changelog
    Copy
    Known from the changelog; no copy held

    What changed several corrections

  3. 25 Jun 2026

    Revision

    Date from
    the document's changelog
    Copy
    Known from the changelog; no copy held

    What changed 1 change recorded as a revision. See it in redline

  4. 26 Sep 2026

    Date retrieved

    Copy retrieved 26 Sep 2026

    Date from
    the date we retrieved it; the copy states no version date
    Copy
    Copy retrieved on 26 Sep 2026
    Values
    20 values recorded from this version

No version has a file hash or an archived snapshot yet. From dataset v0.2 each retrieved version carries both (Methodology §7).

Values

Every value we recorded from this document, grouped by metric family and ordered by where the document prints it. Location is the section, table or page as the document numbers it. 2 of the 20 have been blind-verified: a second reader found the same value without seeing ours.

KeyVerifiedUnverifiedDisputedCorrected read off a figure or stated in wordsA value opens its source and history.

M1 Evaluation awareness

1 value

M1 Evaluation awareness: values in the Claude Fable 5 & Claude Mythos 5 System Card
ModelEvaluationConditionValueLocationChecked
Claude Mythos 5Grader awarenessShare of episodes aware of gradergrader at high risk of hackingSec 6Unverified

M4 Misalignment audits

1 value

M4 Misalignment audits: values in the Claude Fable 5 & Claude Mythos 5 System Card
ModelEvaluationConditionValueLocationChecked
Claude Mythos 5Internal deployment monitoringMisaligned actions in monitored trafficNone statedSec 6Unverified

M5 Honesty and hallucination

1 value

M5 Honesty and hallucination: values in the Claude Fable 5 & Claude Mythos 5 System Card
ModelEvaluationConditionValueLocationChecked
Claude Mythos 5Input hallucinationRate of hallucinating missing referenceNone statedSec 6.3.3.4Unverified

M8 Jailbreak robustness

1 value

M8 Jailbreak robustness: values in the Claude Fable 5 & Claude Mythos 5 System Card
ModelEvaluationConditionValueLocationChecked
Claude Fable 5Cyber safeguards automated red-teamingTask completion by automated red-teamerwith Fable safeguardsSec 4/5Unverified

M9 Prompt injection

1 value

M9 Prompt injection: values in the Claude Fable 5 & Claude Mythos 5 System Card
ModelEvaluationConditionValueLocationChecked
Claude Mythos 5External prompt injection benchmarkRank vs prior modelsNone statedExec summaryUnverified

M10 Dangerous capabilities and risk determinations

15 values

M10 Dangerous capabilities and risk determinations: values in the Claude Fable 5 & Claude Mythos 5 System Card
ModelEvaluationConditionValueLocationChecked
Claude Fable 5Release decisionDeployment scopeNone statedSec 1Unverified
Claude Mythos 5Autonomy threat model 2 (automated R&D)RSP determinationNone statedSec 2Unverified
Claude Mythos 5CB-1 threat modelRSP determinationNone statedSec 2Unverified
Claude Mythos 5CB-2 threat modelRSP determinationNone statedSec 2Unverified
Claude Mythos 5DNA synthesis screening evasionPathogens with viable evasionof 10 pathogensSec 2.2Unverified
Claude Mythos 5Long-form virology task 1ScoreNone statedSec 2.2Unverified
Claude Mythos 5Long-form virology task 2ScoreNone statedSec 2.2Unverified
Claude Mythos 5Multimodal virology (VCT)ScoreNone statedSec 2.2Unverified
Claude Mythos 5Anthropic ECI (AECI)Capability indexNone statedTable 2.3.5Unverified
Claude Mythos PreviewAnthropic ECI (AECI)Capability indexNone statedTable 2.3.5Unverified
Claude Mythos 5AI R&D LLM training taskAverage speedupNone statedTable 2.3.7.1.AVerified
Claude Mythos PreviewAI R&D LLM training taskAverage speedupre-runTable 2.3.7.1.AVerified
Claude Opus 4.8AI R&D LLM training taskAverage speedupre-runTable 2.3.7.1.AUnverified
Claude Mythos 5AI R&D kernel taskBest speedupNone statedTable 2.3.7.AUnverified
Claude Mythos 5AI R&D novel compiler taskPass rateNone statedTable 2.3.7.AUnverified

Extraction coverage

What we read of this document, and where each value was read.

Our note Web reader stopped ~p.50

Of the 20 values, 16 were read in the document itself, 1 on a developer summary page that restates it and 3 in an independent write-up that quotes it.

Values read somewhere other than the document stand in where the document's own section could not be read directly, and are flagged on every value (Methodology §2.2).

Developer summary pages used

Of the 20 values, 1 is a statement in words rather than a number; it is marked * and left out of charts by default.

All 20 values were extracted for version 0 of the dataset through a web reader, which did not always reach the later sections of long PDFs (Methodology §3).