{"schema_version":1,"experiment_id":"eoa_inverse_innovation_exp06_four_proposal_generalization60_20260803","source_assessment_id":"negative_space_design__astronomy_astrophysics:P2:v0","cell_id":"negative_space_design__astronomy_astrophysics","proposal_index":2,"qualification":"EMPIRICAL_PARTNER_CANDIDATE","criteria":{"specific_differentiated_claim":{"status":"YES","reason":"The remaining claim is specific and falsifiable: footprint-local default suppression with an omission inventory, warning exemptions, and exact-registration restoration must outperform both ordinary toggles and synchronized paired views on pixel-grounded morphology judgments without unacceptable safety or time costs."},"credible_problem_signal":{"status":"YES","reason":"Primary research supports the consequential difficulty of faint-structure visual review, and official viewer documentation confirms that overlays routinely occupy astronomical image displays. Direct prevalence and causal harm remain uncertain, but the external signal is credible enough to justify the bounded test."},"identifiable_partner_or_adopter":{"status":"YES","reason":"Rubin/IPAC visualization and data-quality teams are concrete partner classes associated with an existing Firefly-based review surface and capable of authorizing the prototype, defining warning exemptions, granting archived-workflow access, and recruiting qualified reviewers."},"partner_access_is_necessary":{"status":"YES","reason":"Resolving the claim requires representative archived cutouts, overlay geometry and provenance, survey-specific warning classifications, qualified reviewers, and workflow-grounded reference adjudication that ordinary public research cannot supply."},"safe_authorized_first_step":{"status":"YES","reason":"A visualization lead and data-quality lead can authorize a read-only duplicate interface using frozen archived data. The protocol excludes production decisions and requires halting on pixel mismatch, registration drift, undisclosed omission, failed restoration, or suppression of a critical warning."},"bounded_decisive_empirical_design":{"status":"YES","reason":"The preregistered randomized crossover is capped at 60 frozen cutouts, includes the annotated baseline and strongest paired-view rival, records judgment, reliability, warning, contamination, time, and interaction outcomes, and specifies concrete failure thresholds."},"no_material_negative_gate":{"status":"YES","reason":"All verified gates are YES except externally supported problem prevalence, which is UNCERTAIN rather than materially NO and is one of the empirical quantities the proposed partner study directly measures."},"not_merely_more_research":{"status":"YES","reason":"The next action names the authorizers, required dataset and reviewer access, three comparison conditions, immutable-state audits, endpoints, safety margins, and explicit rejection rules."}},"uncertainty_types":["PROBLEM_PREVALENCE","ADOPTER_PULL","INCREMENTAL_EFFECT","WORKFLOW_FIT","DATA_ACCESS"],"partner_profile":"A survey organization operating an overlay-rich astronomical cutout-review workflow, ideally Rubin/IPAC, with a visualization lead authorized to approve a nonoperational prototype, a data-quality lead able to define non-suppressible warnings, access to archived cutouts and annotation provenance, and qualified morphology reviewers.","required_access":"Written prototype approval; up to 60 representative archived cutouts; frozen pixels, stretch, zoom, WCS, masks, footprints, annotations, and provenance; a survey-approved warning taxonomy; qualified reviewers; independently adjudicated reference information where available; and permission to log view condition, toggles, restorations, decisions, confidence, time, and safety errors.","bounded_empirical_test":"Run a preregistered randomized crossover on no more than 60 stratified archived cutouts under three fixed conditions: fully annotated baseline with ordinary toggles, synchronized annotation-free/annotated paired views, and the footprint-local moat with omission inventory and one-action restoration. Audit immutable rendering, registration, disclosure, accessibility, critical-warning visibility, and restoration while recording morphology descriptions, boundary traces, artifact and contamination judgments, classifications, confidence, time, and interactions.","success_condition":"The moat condition outperforms the better comparator on a preregistered primary measure of pixel-grounded correctness or inter-reviewer reliability, while increasing neither missed critical warnings or contamination errors by more than 5 percentage points nor median review time by more than 20%, and while passing every immutable-view, registration, disclosure, restoration, and accessibility audit.","falsification_condition":"Reject the incremental claim if the moat does not outperform the better of ordinary toggles and synchronized paired views on the preregistered primary outcome; if it increases missed critical warnings or contamination errors by more than 5 percentage points; if median review time rises by more than 20%; or if any pixel-identity, registration, omission-disclosure, restoration, or critical-warning audit fails.","rationale":"Adjacent prior art establishes toggles, overlays, and synchronized comparison but does not resolve the proposal's distinct local-suppression claim. The remaining uncertainties are empirical and partner-dependent: overlay-intersection prevalence, incremental reviewer benefit, workflow safety and fit, adopter pull, and access to representative data and reviewers. A concrete survey partner can resolve them through a safe, comparator-based, bounded study."}