Experiment Review and Release Templates — Operating Quantum Computers

This appendix provides copyable templates for experiment admission, release review, and post-run evidence packages.

E.1 Experiment admission template

Illustrative listing · yaml

experiment_admission:
  experiment_id: string
  owner: string
  objective: string
  hypothesis: string
  backend_requirements:
    min_qubits: int
    topology: string
    dynamic_circuits: bool
    analog_mode: bool
  validation_before_qpu:
    exact_simulation_passed: bool
    noisy_simulation_passed: bool
    classical_baseline_id: string
  execution_plan:
    circuits: int
    shots_per_circuit: int
    max_runtime_minutes: int
    reservation_required: bool
  acceptance_criteria:
    metric: string
    threshold: number
  risks:
    - string
  decision:
    status: accepted | rejected | revise
    approver: string

Diagram loads as you read

E.1 Experiment admission template · Figure 1

View diagram source

flowchart TD
    Submit[Submit experiment] --> Validate[Pre-QPU validation]
    Validate --> Budget[Budget review]
    Budget --> Risk[Risk review]
    Risk --> Decision{Admit?}
    Decision -- yes --> Queue[Queue or reserve]
    Decision -- no --> Revise[Revise]

E.2 Release review template

Illustrative listing · yaml

workflow_release_review:
  workflow_id: string
  version: string
  owner: string
  release_type: alpha | beta | production | retirement
  evidence:
    known_answer_tests: string
    baseline_comparison: string
    noisy_regression: string
    hardware_trial: string
    cost_model: string
    security_review: string
  service_contract:
    result_schema_version: string
    uncertainty_required: bool
    provenance_required: bool
  rollback_plan: string
  decision:
    approved: bool
    approvers: [string]
    conditions: [string]

Diagram loads as you read

E.2 Release review template · Figure 2

View diagram source

flowchart LR
    Correctness[Correctness] --> Release[Release decision]
    Baseline[Baseline] --> Release
    Cost[Cost model] --> Release
    Security[Security] --> Release
    Rollback[Rollback plan] --> Release

E.3 Post-run evidence package

Illustrative listing · yaml

post_run_evidence:
  run_id: string
  workload_id: string
  backend_id: string
  execution_window: string
  artifacts:
    raw_shots: uri
    compiled_circuits: uri
    calibration_snapshot: uri
    logs: uri
    result_summary: uri
  interpretation:
    primary_metric: string
    uncertainty: string
    baseline_delta: string
    anomalies: [string]
  disposition:
    accepted: bool
    quarantined: bool
    follow_up_actions: [string]

Diagram loads as you read

E.3 Post-run evidence package · Figure 3

View diagram source

flowchart TB
    Run[Run] --> Artifacts[Artifacts]
    Artifacts --> Interpret[Interpretation]
    Interpret --> Disposition[Disposition]
    Disposition --> Archive[Archive]
    Disposition --> Followup[Follow-up]

E.4 Review questions

Use these prompts in every review:

What decision does this experiment support?
What is the classical baseline?
What uncertainty is acceptable?
What hardware state is required?
What could make the result invalid?
What will be archived?
Who owns follow-up?