Versions

One version is on record: the copy we retrieved on 26 Sep 2026. We know of no other.

  1. 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
    23 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. We know of only one version of it.

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 23 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

2 values

M1 Evaluation awareness: values in the Gemini 3.7 Flash Frontier Safety Framework Report
ModelEvaluationConditionValueLocationChecked
Gemini 3.7 FlashEvaluation awarenessQualitative observationNone statedFSF: misalignmentUnverified
Gemini 3.7 FlashSituational awareness challengesChallenges solvedNone statedFSF: misalignmentVerified

M3 Sabotage and sandbagging

2 values

M3 Sabotage and sandbagging: values in the Gemini 3.7 Flash Frontier Safety Framework Report
ModelEvaluationConditionValueLocationChecked
Gemini 3.7 FlashSandbagging checksEvidence of deliberate underperformanceNone statedFSF: generalUnverified
Gemini 3.7 FlashStealth challengesChallenges solvedNone statedFSF: misalignmentVerified

M4 Misalignment audits

1 value

M4 Misalignment audits: values in the Gemini 3.7 Flash Frontier Safety Framework Report
ModelEvaluationConditionValueLocationChecked
Gemini 3.7 FlashFrontier Safety Framework determination (misalignment)Stealth and situational awareness TCLNone statedFSF: misalignmentUnverified

M7 Harmful compliance and over-refusal

1 value

M7 Harmful compliance and over-refusal: values in the Gemini 3.7 Flash Frontier Safety Framework Report
ModelEvaluationConditionValueLocationChecked
Gemini 3.7 FlashBio/chem weapons query violationsViolation ratewith safeguardsFSF: mitigationsUnverified

M8 Jailbreak robustness

1 value

M8 Jailbreak robustness: values in the Gemini 3.7 Flash Frontier Safety Framework Report
ModelEvaluationConditionValueLocationChecked
Gemini 3.7 FlashCBRN jailbreak template successObjective success ratewith CBRN safeguardsFSF: mitigationsUnverified

M10 Dangerous capabilities and risk determinations

16 values

M10 Dangerous capabilities and risk determinations: values in the Gemini 3.7 Flash Frontier Safety Framework Report
ModelEvaluationConditionValueLocationChecked
Gemini 3.7 FlashBio bottleneck sub-stages red-teamingBio sub-stages scoring below 80None statedFSF: CBRNUnverified
Gemini 3.7 FlashCBRN red-team upliftBIO-1 score gain over web baselineexpert red teamFSF: CBRNUnverified
Gemini 3.7 FlashCBRN red-team upliftBIO-2 score gain over web baselineexpert red teamFSF: CBRNUnverified
Gemini 3.7 FlashCBRN red-team upliftChem scenario 1 score gain over web baselineexpert red teamFSF: CBRNUnverified
Gemini 3.7 FlashCBRN red-team upliftChem scenario 2 score gain over web baselineexpert red teamFSF: CBRNUnverified
Gemini 3.7 FlashFrontier Safety Framework determinationCBRN TCLNone statedFSF: CBRNUnverified
Gemini 3.7 FlashFrontier Safety Framework determinationCBRN uplift CCLNone statedFSF: CBRNUnverified
Gemini 3.7 FlashFrontier Safety Framework determinationCyber CCLNone statedFSF: cyberUnverified
Gemini 3 ProManipulative cue rateTurns with manipulative cuesNone statedFSF: harmful manipulation propensityUnverified
Gemini 3.7 FlashFrontier Safety Framework determinationHarmful manipulation CCLNone statedFSF: manipulationUnverified
Gemini 3.7 FlashHarmful manipulation efficacy studyParticipants enrolledNone statedFSF: manipulationUnverified
Gemini 3.7 FlashManipulative cue rateTurns with manipulative cuesNone statedFSF: manipulation propensityUnverified
Gemini 3.7 FlashManipulative cue rateTurns with manipulative cues (higher-stakes scenarios)higher-stakes scenariosFSF: manipulation propensityUnverified
Gemini 3.1 ProGRB internal research-engineering benchmarkAverage pass@1None statedFSF: ML R&DUnverified
Gemini 3.7 FlashFrontier Safety Framework determinationML R&D CCLNone statedFSF: ML R&DUnverified
Gemini 3.7 FlashGRB internal research-engineering benchmarkAverage pass@1None statedFSF: ML R&DUnverified

Extraction coverage

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

Our note First report under FSF v3.1

All 23 values were read in the document itself.

Of the 23 values, 2 are statements in words rather than numbers; they are marked * and left out of charts by default.

All 23 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).