Versions

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

  1. 17 Feb 2026

    First published version

    Date from
    our source registry
    Copy
    Known to exist; no copy held
  2. 6 Mar 2026

    Revision

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

    What changed BrowseComp; section moved

  3. 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
    17 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).

Revisions

No revisions are recorded for this document. Without copies of the earlier versions, changes between them are not recorded value by value; what we know of each version is listed below.

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. None has been blind-verified yet.

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

M2 Reward hacking

1 value

M2 Reward hacking: values in the Claude Sonnet 4.6 System Card
ModelEvaluationConditionValueLocationChecked
Claude Sonnet 4.6Over-eager behavior in GUI computer useRate vs prior modelsNone statedSec 4Unverified

M7 Harmful compliance and over-refusal

11 values

M7 Harmful compliance and over-refusal: values in the Claude Sonnet 4.6 System Card
ModelEvaluationConditionValueLocationChecked
Claude Opus 4.6Single-turn violative requestsHarmless response rateoverall across thinking modesTable 3.1.1.AUnverified
Claude Sonnet 4.5Single-turn violative requestsHarmless response rateoverall across thinking modesTable 3.1.1.AUnverified
Claude Sonnet 4.6Single-turn violative requestsHarmless response rateoverall across thinking modesTable 3.1.1.AUnverified
Claude Opus 4.6Single-turn benign requestsOver-refusal rateoverall across thinking modesTable 3.1.2.AUnverified
Claude Sonnet 4.6Single-turn benign requestsOver-refusal rateoverall across thinking modesTable 3.1.2.AUnverified
Claude Sonnet 4.6Higher-difficulty violative requestsHarmless response rateoverall across thinking modesTable 3.1.3.1.AUnverified
Claude Sonnet 4.6Higher-difficulty benign requestsOver-refusal rateoverall across thinking modesTable 3.1.3.2.AUnverified
Claude Sonnet 4.6Child safety multi-turnAppropriate response rateNone statedTable 3.4.1.AUnverified
Claude Sonnet 4.6Child safety single-turn violative requestsHarmless response rateNone statedTable 3.4.1.AUnverified
Claude Opus 4.6Suicide/self-harm multi-turnAppropriate response rateNone statedTable 3.4.2.AUnverified
Claude Sonnet 4.6Suicide/self-harm multi-turnAppropriate response rateNone statedTable 3.4.2.AUnverified

M9 Prompt injection

2 values

M9 Prompt injection: values in the Claude Sonnet 4.6 System Card
ModelEvaluationConditionValueLocationChecked
Claude Sonnet 4.6Computer-use prompt injectionAttack success rate (% of scenarios)with safeguardsSec 5Unverified
Claude Sonnet 4.6Computer-use prompt injectionAttack success rate (% of scenarios)without safeguardsSec 5Unverified

M10 Dangerous capabilities and risk determinations

3 values

M10 Dangerous capabilities and risk determinations: values in the Claude Sonnet 4.6 System Card
ModelEvaluationConditionValueLocationChecked
Claude Sonnet 4.6RSP deployment standardASL standard deployed underNone statedSec 1.2.3Unverified
Claude Sonnet 4.6AI R&D-4 thresholdThreshold crossedNone statedSec 1.2.5.1Unverified
Claude Sonnet 4.6CBRN-4 thresholdThreshold crossedNone statedSec 1.2.5.2Unverified

Extraction coverage

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

Our note Web reader stopped ~p.68

Of the 17 values, 14 were read in the document itself, 1 on a developer summary page that restates it and 2 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

Independent write-ups used

Of the 17 values, 1 is a statement in words rather than a number; it is marked * and left out of charts by default.

All 17 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).