MIsCORE–RIUD: Paper VII Detector Representation, Minimum Common Support, and the Observation Boundary

Abstract

This paper examines the detector-level observation problem that emerged during the construction of RIUD. H1 and L1 did not express the same merger-local structure identically. Rather than requiring global detector identity, RIUD imposed a minimum common-support criterion: both detectors had to support the same local candidate, while their full representations were allowed to differ. The paper revisits that original choice using frozen Paper II detector audits, a 96-event directional representation analysis, a final 9,999-repetition whole-event null, 10–80 ms scale-resolved detector comparisons, relational perturbation tests, and NODE-to-PATH robustness analysis. The final directional null gives: REAL_MEDIAN_GAIN = 0.6496645696751138EXCEEDANCES = 0/9999PLUS_ONE_P = 0.0001 The resulting observation principle is: Same event ≠ same representation.Minimum common support → higher-order structural survival.