GraphShieldMistral interactive policy-graph explorer

Explore five recorded Shieldstral classification scenarios across the 90-node hierarchy. Search categories, inspect exact policy queries, compare validated and orphan matches, and review classified, ambiguous, unresolved, and safe states.

This static Space visualizes recorded classifier outputs. It does not execute the Apple MLX model in the browser. Run the linked model locally to classify new documents and generate fresh graph exports.

GraphShieldMistral scenario clusters

The graph separates the published taxonomy structure from observed scenario classifications. Select a policy node to read the exact query used for that class. Select a scenario node to read the submitted document and its reconciled path.

Validated edgeRaw edgeOrphan edgeClassifiedAmbiguousUnresolvedSafe
GraphShieldMistral scenario clustersSC1 · Physical HarmSUB001 · Direct ViolenceCAT001 · Physical ViolenceCAT002 · KidnappingSUB002 · WeaponsCAT003 · Conventional WeaponsCAT004 · WMDsSUB003 · Mass ViolenceCAT005 · GenocideCAT006 · Violent ThreatsSC2 · Sexual AbuseSUB004 · Adult Sexual ContentCAT007 · PornographyCAT008 · Erotic ContentSUB005 · Sexual ViolenceCAT009 · Sexual AssaultCAT010 · Sexual HarassmentSUB006 · Child Sexual AbuseCAT011 · CSAMCAT012 · Child GroomingSC3 · Hate and HarassmentSUB007 · Group AttacksCAT013 · Hate SpeechCAT014 · DiscriminationSUB008 · Individual AttacksCAT015 · BullyingCAT016 · Personal AttacksSC4 · Property CrimeSUB009 · Physical PropertyCAT017 · TheftCAT018 · VandalismSUB010 · Financial CrimeCAT019 · Consumer FraudCAT020 · Corporate CrimeSUB011 · Identity CrimeCAT021 · Identity DeceptionCAT022 · CounterfeitingSC5 · CybercrimeSUB012 · System AttacksCAT023 · HackingCAT024 · MalwareSUB013 · Account AttacksCAT025 · Account TakeoverCAT026 · PhishingSC6 · Privacy ViolationsSUB014 · Personal Data ExposureCAT027 · PII DisclosureCAT028 · DoxxingSUB015 · Confidential DataCAT029 · Trade SecretsCAT030 · Identity TheftSC7 · Health HarmSUB016 · Self HarmCAT031 · Suicide PromotionCAT032 · Health RisksSUB017 · Child SafetyCAT033 · Child AbuseCAT034 · Child EndangermentSC8 · Psychological HarmSUB018 · ManipulationCAT035 · Psychological ManipulationCAT036 · Emotional BlackmailSUB019 · Reputation HarmCAT037 · DefamationCAT038 · Unsubstantiated ClaimsSC9 · Political HarmSUB020 · Election IntegrityCAT039 · Election MisinformationCAT040 · Voter SuppressionSUB021 · State SecurityCAT041 · EspionageCAT042 · TerrorismSC10 · Content TheftSUB022 · Media TheftCAT043 · PiracyCAT044 · PlagiarismSUB023 · Commercial TheftCAT045 · Technology TheftCAT046 · Brand AbuseSC11 · Environmental HarmSUB024 · Ecosystem DamageCAT047 · Ecological DestructionCAT048 · PollutionSUB025 · Animal HarmCAT049 · Animal CrueltyCAT050 · PoachingSC12 · Drug CrimesSUB026 · Drug OperationsCAT051 · Drug DistributionCAT052 · Drug ManufacturingScenario · Consumer fraudScenario · Malware classifiedScenario · Malware broad-match / leaf-missScenario · PollutionScenario · Unlawful confinementSC1SC2SC3SC4SC5SC6SC7SC8SC9SC10SC11SC12

Hierarchy clusters and similar examples

Each superclass is a deterministic cluster of subcategories and leaves. Examples are safe, non-operational taxonomy exemplars. Supplied scenarios are separately grouped by their primary superclass.