How do we know a creative improvement is not just a searched winner or model preference?
Separate richer perception, evidence-grounded meaning, real encounter, typed consequence, search history, and independent holdout before promoting a creative prior.
01 / Current position
A hypothesis is not the current judgment.
The dossier preserves the live research position. Dated articles preserve the complete evidence and argument that changed it.
Exact artifact identity, search-aware evidence, realized encounter, grounded consequence, and independent OOS support can support bounded creative learning without a universal reward model.
Testing after R4–R6. Rich perception earned medium-local equipment, grounding proved more stable than fine labels, and R6 showed a 5,000-candidate zero-effect search could manufacture an impressive winner that failed stronger evidence. One reveal-order effect replicated directionally on pristine content for one Agent-observer class.
02 / Decision boundary
What keeps this Question alive?
Agent-scale generation makes false winners cheap: a system can search until one variant looks exceptional and then mistake selection noise or one observer class for creative truth.
Change observer class/provider and repeat frozen interventions across new artifact worlds. Introduce human evidence only when making a human comprehension, preference, trust, memory, or behavior claim.
A simpler production/evaluation loop reaches the same grounded and OOS conclusions without preserving search history or encounter identity, or the retained creative effect disappears consistently under independent observers and content.
03 / Supporting publications
Complete arguments connected to this Question.
1 dated publication currently document this research line.
04 / Source discipline
The dossier is an index, not the evidence authority.
Owns the current judgment, next test, and deletion condition.
Own complete dated arguments, limitations, comparisons, and source links.
Own exact code, tests, releases, receipts, and machine evidence.