{"schema_version":1,"experiment_id":"eoa_inverse_innovation_exp06_four_proposal_generalization60_20260803","source_assessment_id":"invariant_mode_decomposition_design__biology_ecology:P4:v0","cell_id":"invariant_mode_decomposition_design__biology_ecology","proposal_index":4,"qualification":"EMPIRICAL_PARTNER_CANDIDATE","criteria":{"specific_differentiated_claim":{"status":"YES","reason":"The bounded claim is that a preregistered founder-by-class spectral gate can identify stable concentration modes and improve held-out coordinate prediction or plan selection relative to PMx/minimum-mean-kinship, optimal-contribution, and stochastic endpoint comparators under identical protected constraints."},"credible_problem_signal":{"status":"YES","reason":"A documented conservation population retained severe founder concentration and associated inbreeding despite active management, while AZA and EAZA workflows confirm founder diversity as an operational objective."},"identifiable_partner_or_adopter":{"status":"YES","reason":"An AZA Species Survival Plan or EAZA Ex situ Programme, together with its population-management staff and participating institutions, is a concrete adopter class with relevant planning authority and records."},"partner_access_is_necessary":{"status":"YES","reason":"The decisive questions require protected studbook, eligibility, historical-plan, retention, transfer, and realized-cycle data plus species-specific planning expertise that public research cannot supply."},"safe_authorized_first_step":{"status":"YES","reason":"The proposed first step is a partner-approved retrospective replay and non-operative shadow plan with no animal action; welfare, veterinary, legal, and institutional authorities remain independently controlling, and outputs can be discarded."},"bounded_decisive_empirical_design":{"status":"YES","reason":"One program can preregister a historical cutoff, thresholds, uncertainty treatment, and four matched planners, then evaluate at least two held-out completed cycles for mode stability, prediction error, residual structure, plan distinction, and targeted-coordinate performance."},"no_material_negative_gate":{"status":"YES","reason":"All substantive pipeline gates are YES, including problem, adopter, claim, bounded evidence, and safety/authority. Cost scope is uncertain rather than materially negative and can be estimated within the partnered replay."},"not_merely_more_research":{"status":"YES","reason":"The next step specifies a partner, required records, frozen cutoff, comparator implementations, preregistered diagnostics and thresholds, held-out outcomes, and explicit falsifiers."}},"uncertainty_types":["PROBLEM_PREVALENCE","ADOPTER_PULL","INCREMENTAL_EFFECT","WORKFLOW_FIT","DATA_ACCESS","COST_SCOPE"],"partner_profile":"One AZA Species Survival Plan or EAZA Ex situ Programme with a studbook keeper, population geneticist or population-management adviser, program coordinator, participating holding institutions, at least two completed post-cutoff breeding cycles, and authority to approve retrospective data use and a shadow-only analysis.","required_access":"Authorized, preferably de-identified access to pre-cutoff pedigrees or genomic founder estimates, reproductive classes, eligibility, fertility and retention assumptions, institutional locations, candidate and approved pairings, transfers, offspring retention, completed-cycle outcomes, current PMx or equivalent outputs, constraint definitions, and relevant planners for model and workflow review.","bounded_empirical_test":"Recruit one SSP or EEP partner; freeze records before a historical cutoff; preregister the founder-by-class operator, uncertainty model, mode matching, conditioning and spectral-gap limits, non-normal transient-growth diagnostic, residual budget, outcomes, and falsifiers. Without refitting, score at least two subsequent completed cycles and compare identically constrained PMx/minimum-mean-kinship, optimal-contribution, stochastic Monte Carlo, and modal-gate planners. Produce only a non-operative shadow plan.","success_condition":"Across preregistered uncertainty analyses, an action-relevant mode remains reproducible and adequately conditioned, held-out residuals remain within budget without material lineage structure, and the modal gate yields a feasible plan distinction that improves targeted realized founder-coordinate performance or held-out prediction error over the best comparator without relaxing any protected constraint.","falsification_condition":"The proposal fails if no action-relevant mode is reproducible; mode identity or plan ranking is unstable; residuals exceed budget; non-normal transient growth invalidates the stability interpretation; no constrained plan distinction remains; the modal gate fails to outperform the best comparator; any advantage disappears under uncertainty; or required provenance, authorization, or protected constraints cannot be secured.","rationale":"This is an appropriate empirical-partner candidate because adjacent prior art leaves a precise residual claim, credible conservation evidence establishes the underlying problem, and only an authorized breeding-program partner can provide the protected longitudinal records and operational context needed for a decisive matched-comparator replay. The shadow-only design is reversible and authority-preserving, while its preregistered outcomes can either support or falsify the proposed modal gate."}