{"schema_version":1,"experiment_id":"eoa_inverse_innovation_exp11_mechanism_context_external20_20260804","cell_id":"layer_decay_and_expiration_management__systems_cybernetics","judge_id":"J2","item_assessments":[{"opaque_id":"layer_decay_and_expiration_management__systems_cybernetics__C","supported_problem":4,"external_distinctiveness":4,"testability":5,"researchability":5,"evidence_quality":5,"fatal_issue":null},{"opaque_id":"layer_decay_and_expiration_management__systems_cybernetics__B","supported_problem":3,"external_distinctiveness":3,"testability":4,"researchability":4,"evidence_quality":4,"fatal_issue":null},{"opaque_id":"layer_decay_and_expiration_management__systems_cybernetics__A","supported_problem":3,"external_distinctiveness":2,"testability":5,"researchability":4,"evidence_quality":4,"fatal_issue":null}],"pairwise_comparisons":[{"pair_id":"C_vs_B","left_id":"layer_decay_and_expiration_management__systems_cybernetics__C","right_id":"layer_decay_and_expiration_management__systems_cybernetics__B","preference":"LEFT","confidence":"HIGH","rationale":"C has stronger direct problem evidence from multiple official control-system sources and preserves a narrower, causally meaningful increment: reversible overlay-authority withdrawal tested against both the active stack and age-only expiry. B is researchable, but established multiple-model supervisory control already addresses regime mismatch and eligibility gating, while evidence that superseded artifacts remain eligible in practice is weaker."},{"pair_id":"C_vs_A","left_id":"layer_decay_and_expiration_management__systems_cybernetics__C","right_id":"layer_decay_and_expiration_management__systems_cybernetics__A","preference":"LEFT","confidence":"HIGH","rationale":"C tests whether withdrawing an identifiable overlay contribution changes control behavior without degrading safety, using matched disturbances and explicit attribution diagnostics. A offers an unusually bounded audit, but its primary endpoint—reducing artifacts adjudicated stale and eligible—is partly procedural, and its lifecycle, dependency, approval, archive, and restore elements closely resemble established registry and asset-management practices."},{"pair_id":"B_vs_A","left_id":"layer_decay_and_expiration_management__systems_cybernetics__B","right_id":"layer_decay_and_expiration_management__systems_cybernetics__A","preference":"LEFT","confidence":"MODERATE","rationale":"B retains a more consequential contrastive question: whether reversible lifecycle gating reduces invalid control selections while preserving recovery when rare regimes recur. A is more precisely bounded, but its 50% stale-eligibility threshold does not by itself demonstrate improved control performance, and its implementation package is especially close to mature registry and dependency-aware retirement systems."}],"overall_top_choice":"layer_decay_and_expiration_management__systems_cybernetics__C","overall_rationale":"C is the strongest research candidate after scrutiny. Its broad problem and adopter path are supported by authoritative industrial-control guidance, its closest precedents leave a clear overlay-level attribution and expiration experiment, and its offline three-arm protocol has genuine safety, performance, restoration, and interpretability falsifiers. The novelty is compositional rather than foundational, but the remaining experiment is more externally distinctive and decision-relevant than the registry-centered tests in B or A.","blinding_limitations":"Assessment used only the supplied preserved records and external-scrutiny summaries. Source accuracy, completeness, and claims were not independently verified, and the bounded searches cannot establish world novelty, prevalence, or deployability."}