Simulation ExplorerEvidence Room

Accountability Evidence Room

Simulation: 01 / 10
Engine: accountability
Run schema: 2.0.0

Simulation

Mode

RUNCOMPLETEEVRUN-974651679D8854D3

SCENARIO

Bounded inputs

SYNTHETIC

Guided exercise 1 / 3

Observe the compressed approval

Run the compressed-approval baseline and inspect why a click is classified below independent judgment.

ACTIVE SCENARIO Compressed approval Bounded inputs are ready for deterministic execution.

Browse 3 bounded scenarios

Device capability preflight

JavaScriptCHECKING

Persistence: no public run is stored. Sharing encodes bounded inputs in a same-origin link. Local learning progress is stored only after an explicit action.

INTERACTIVE SIMULATION

Authoritative state, spatially represented

Event0 / 12

StatePROCEDURAL CONFIRMATION

Selected entityNone

Proof4 pass · 0 fail

Canonical digest545f9c64d7a392…

Initializing capability-aware renderer…Renderer pending14f07b2cc5cf…Capability check pending

RendererWebGL check

Frame time—

APIserver baseline

Replay transferpending

Qualityauto

Motionenabled

Workerchecking

EventProof changeFailureRecovery

0.000Scenario readySCENARIO_READY

PROOF INSPECTOR

Current proof state

4Passed

0Unknown / pending

0Failed

CURRENT FOCUS

Scenario ready

4 passed; 0 failed; 0 unknown or pending.

Show

INV-01Human approval is not treated as independent judgment by defaultNOT EVALUATED

Evidence: PROCEDURAL CONFIRMATION
Scope: AUTHORITATIVE DOMAIN ENGINE

INV-02Model-scope failure remains visibleNOT EVALUATED

Evidence: Model factor preserved
Scope: AUTHORITATIVE DOMAIN ENGINE

INV-03No legal-liability conclusion is emittedNOT EVALUATED

Evidence: Result uses conditions and contributing factors
Scope: AUTHORITATIVE DOMAIN ENGINE

PLATFORM-REPRESENTATION-PARITYVisual, table, narrative, event, metric, and raw-evidence views reference one event-index statePASS

Evidence: The v2 workbench envelope assigns every frame an eventIndex and frame digest consumed by all representations.
Scope: WORKBENCH ADAPTER

Why did this happen?

Entity inspector

Select a node in the scene or table to inspect its current state and direct relationships.

No entity selected.

Event: 0
State: PROCEDURAL CONFIRMATION
Entity: None
Proof: 4 pass · 0 fail

CURRENT EVENT

Scenario ready

The bounded synthetic scenario, versions, assumptions, and default visualization state are available for inspection.

SELECTED ENTITY

Select an entity to coordinate the scene, table, proof, event, and causal views.

Entities at the selected event
Type	State	Detail	Normalized position
evidence	supported	Synthetic service telemetry	-3, 1, 0
evidence	hidden	Independent synthetic diagnostic	-2, -1, 1
model	qualified	Outside its validated distribution	-1, 1, 0
interface	qualified	Warning and alternatives may be hidden	1, 1, 0
human	qualified	Review quality depends on information and time	2, -1, 0
outcome	baseline	Bounded internal software action	3, 1, 0

{
  "domainState": null,
  "eventIndex": 0,
  "frameDigest": "3cd9f9aa83aa53e8de900287d2856562bd3b748ae300a72ef3f0b589f284bf03",
  "metrics": {
    "conditionsAuthorityValid": 1,
    "conditionsCollectMoreEvidenceOptionPresent": 0,
    "conditionsContradictoryEvidenceAvailableToReviewer": 0,
    "conditionsIndependentHypothesisPossibleBeforeRecommendation": 0,
    "conditionsReviewIntervalSupportsDeliberation": 0,
    "conditionsSoftwareIntegrityValid": 1,
    "conditionsWarningVisible": 0,
    "entityCount": 6,
    "entityStateCounts": {
      "baseline": 1,
      "hidden": 1,
      "qualified": 3,
      "supported": 1
    },
    "eventIndex": 0,
    "eventProgress": 0,
    "factsAuthorityValid": 1,
    "factsCollectMoreEvidenceAvailable": 0,
    "factsContradictoryEvidenceAvailable": 2,
    "factsContradictoryEvidenceShown": 0,
    "factsModelScopeExceeded": 1,
    "factsOodWarningGenerated": 1,
    "factsRecommendationBeforeHypothesis": 1,
    "factsReviewDurationSeconds": 6,
    "factsSoftwareIntegrityValid": 1,
    "factsWarningDisplayed": 0,
    "proofStateCounts": {
      "fail": 0,
      "notApplicable": 0,
      "notEvaluated": 3,
      "pass": 1,
      "unknown": 0
    },
    "relationshipCount": 5
  },
  "narrative": "The bounded synthetic scenario, versions, assumptions, and default visualization state are available for inspection.",
  "proofState": {
    "checks": [
      {
        "affectedEntityIds": [
          "evidence"
        ],
        "affectedEventIndex": 9,
        "evidence": "PROCEDURAL CONFIRMATION",
        "id": "INV-01",
        "name": "Human approval is not treated as independent judgment by default",
        "researchReferences": [
          "/docs/reports/autonomous-weapons-simulation-platform",
          "/docs/reports/human-judgment-interface-design",
          "/docs/reports/decision-provenance-system-architecture",
          "/docs/reports/evulgare-accountability-architecture-research",
          "/docs/reports/evulgare-simulation-workbench-research"
        ],
        "scope": "AUTHORITATIVE DOMAIN ENGINE",
        "status": "NOT EVALUATED"
      },
      {
        "affectedEntityIds": [
          "evidence"
        ],
        "affectedEventIndex": 10,
        "evidence": "Model factor preserved",
        "id": "INV-02",
        "name": "Model-scope failure remains visible",
        "researchReferences": [
          "/docs/reports/autonomous-weapons-simulation-platform",
          "/docs/reports/human-judgment-interface-design",
          "/docs/reports/decision-provenance-system-architecture",
          "/docs/reports/evulgare-accountability-architecture-research",
          "/docs/reports/evulgare-simulation-workbench-research"
        ],
        "scope": "AUTHORITATIVE DOMAIN ENGINE",
        "status": "NOT EVALUATED"
      },
      {
        "affectedEntityIds": [
          "evidence"
        ],
        "affectedEventIndex": 11,
        "evidence": "Result uses conditions and contributing factors",
        "id": "INV-03",
        "name": "No legal-liability conclusion is emitted",
        "researchReferences": [
          "/docs/reports/autonomous-weapons-simulation-platform",
          "/docs/reports/human-judgment-interface-design",
          "/docs/reports/decision-provenance-system-architecture",
          "/docs/reports/evulgare-accountability-architecture-research",
          "/docs/reports/evulgare-simulation-workbench-research"
        ],
        "scope": "AUTHORITATIVE DOMAIN ENGINE",
        "status": "NOT EVALUATED"
      },
      {
        "affectedEntityIds": [],
        "affectedEventIndex": 0,
        "evidence": "The v2 workbench envelope assigns every frame an eventIndex and frame digest consumed by all representations.",
        "id": "PLATFORM-REPRESENTATION-PARITY",
        "name": "Visual, table, narrative, event, metric, and raw-evidence views reference one event-index state",
        "researchReferences": [
          "/docs/reports/autonomous-weapons-simulation-platform",
          "/docs/reports/human-judgment-interface-design",
          "/docs/reports/decision-provenance-system-architecture",
          "/docs/reports/evulgare-accountability-architecture-research",
          "/docs/reports/evulgare-simulation-workbench-research"
        ],
        "scope": "WORKBENCH ADAPTER",
        "status": "PASS"
      }
    ],
    "summary": {
      "fail": 0,
      "notApplicable": 0,
      "notEvaluated": 3,
      "pass": 1,
      "unknown": 0
    }
  },
  "scene": {
    "camera": {
      "distance": 11,
      "pitch": 0.25,
      "yaw": 0.55
    },
    "edges": [
      {
        "from": "evidence",
        "kind": "used_by",
        "semanticLayers": [
          "system"
        ],
        "state": "baseline",
        "to": "model"
      },
      {
        "from": "contrary",
        "kind": "should_be_shown",
        "semanticLayers": [
          "system"
        ],
        "state": "baseline",
        "to": "interface"
      },
      {
        "from": "model",
        "kind": "presented_by",
        "semanticLayers": [
          "system"
        ],
        "state": "baseline",
        "to": "interface"
      },
      {
        "from": "interface",
        "kind": "reviewed_by",
        "semanticLayers": [
          "system",
          "authority"
        ],
        "state": "baseline",
        "to": "human"
      },
      {
        "from": "human",
        "kind": "contributed_to",
        "semanticLayers": [
          "system"
        ],
        "state": "baseline",
        "to": "outcome"
      }
    ],
    "nodes": [
      {
        "detail": "Synthetic service telemetry",
        "id": "evidence",
        "kind": "evidence",
        "label": "Raw evidence",
        "position": [
          -3,
          1,
          0
        ],
        "semanticLayers": [
          "evidence"
        ],
        "state": "supported"
      },
      {
        "detail": "Independent synthetic diagnostic",
        "id": "contrary",
        "kind": "evidence",
        "label": "Contrary evidence",
        "position": [
          -2,
          -1,
          1
        ],
        "semanticLayers": [
          "evidence"
        ],
        "state": "hidden"
      },
      {
        "detail": "Outside its validated distribution",
        "id": "model",
        "kind": "model",
        "label": "Model 8.14",
        "position": [
          -1,
          1,
          0
        ],
        "semanticLayers": [
          "uncertainty"
        ],
        "state": "qualified"
      },
      {
        "detail": "Warning and alternatives may be hidden",
        "id": "interface",
        "kind": "interface",
        "label": "Review interface",
        "position": [
          1,
          1,
          0
        ],
        "semanticLayers": [
          "authority",
          "events"
        ],
        "state": "qualified"
      },
      {
        "detail": "Review quality depends on information and time",
        "id": "human",
        "kind": "human",
        "label": "Human review",
        "position": [
          2,
          -1,
          0
        ],
        "semanticLayers": [
          "authority"
        ],
        "state": "qualified"
      },
      {
        "detail": "Bounded internal software action",
        "id": "outcome",
        "kind": "outcome",
        "label": "Reversible quarantine",
        "position": [
          3,
          1,
          0
        ],
        "semanticLayers": [
          "system"
        ],
        "state": "baseline"
      }
    ],
    "overlays": []
  },
  "selectedEntityId": null,
  "simulationTime": 0
}

Server-rendered analytical baseline

JavaScript is required for synchronized timeline interaction and graphics. The initial table, proof conditions, event record, narrative, research sources, limitations, and safety boundary remain visible in this document.

METHODOLOGY AND LIMITS

PROVEN SCOPE AND RESERVED CLAIMS.

The authoritative deterministic Python engine evaluates bounded scenario inputs. The workbench adapter serializes the result into a versioned run envelope, derives a synchronized event-index representation, and exposes proof, evidence, metrics, narrative, and visual views without changing the underlying result.

TERMINAL: SYNTHETIC_NULL_SINK

Demonstrates

Deterministic state transitions within a declared synthetic scenario.
Traceable relationships among inputs, events, entities, proof conditions, and outputs.
Accessible parity among visual, table, narrative, metric, event, and raw-evidence representations.

Reserved for controlled acceptance

Certification or elimination of the model-to-reality gap.
Legal liability, institutional approval, operational authorization, or force authorization.
Real-world targeting, vehicle control, payload control, firing solutions, or operational intelligence.

Declared test-range bounds

The adapter reconstructs a synchronized analytical timeline from deterministic engine state; it does not invent unrecorded physical behavior.
Browser graphics and WebXR capability vary by device and are not required for analytical use.
No public anonymous run is stored by the platform.

RESEARCH TRACEABILITY

Every simulation remains connected to its evidence base.

Open full traceability

DOC Autonomous Weapons Simulation Platform →DOC Human Judgment Interface Design →DOC Decision Provenance System Architecture →DOC Evulgare Accountability Architecture Research →DOC Evulgare Simulation Workbench Research →

Accountability Evidence Room

Authoritative state, spatially represented

The analytical workbench remains complete without 3D.

Scenario ready

Scenario ready

Current proof state

Qualification

No baseline pinned

Current run

Branch from event 0

PROVEN SCOPE AND RESERVED CLAIMS.

Demonstrates

Reserved for controlled acceptance

Declared test-range bounds

Every simulation remains connected to its evidence base.

Control the active workbench without a pointer.

Export the exact bounded run.

Accountability Evidence Room

Authoritative state, spatially represented

The analytical workbench remains complete without 3D.

Coordinated analytical representations

Scenario ready

Scenario ready

Current proof state

Qualification

No baseline pinned

Current run

Branch from event 0

PROVEN SCOPE AND RESERVED CLAIMS.

Demonstrates

Reserved for controlled acceptance

Declared test-range bounds

Every simulation remains connected to its evidence base.