nx_paper_panel.nx

nx_paper_panel.nx -- ADMISSIBILITY of a manuscript grade (AS-2). ★WHY THIS IS A COMPOSITION AND NOT A NEW ORGAN. The obvious move after shipping a manuscript grader (nx_paperbench) was to build "nx_paper_referee". A check-before-build sweep found the ecosystem ALREADY HAS the referee machinery, in three organs: nx_referee.nx ref_no_self_grade -- the scorer must not be a competitor nx_referee_v2.nx rf2_judge_independent (R6 preference-leakage), rf2_holdout_clean (R7 contamination) nx_peer_review.nx pr_agreement / pr_resolve / pr_divergence_cause -- dual grade + tie-break Building a fourth would have been duplication. This organ contributes the ONE thing that did not exist: binding those gates to a MANUSCRIPT grade so that a score is inadmissible unless its grader is independent of the paper's author. ★★AND IT INDICTS OUR OWN PUBLISHED WORK. RT-004 scored 1000 permille on nx_paperbench -- but that grader was written by the same author, in the same session, as the paper it graded. rf2_judge_independent has said since it was written that such a judge is INVALID. So the honest verdict on RT-004's headline is PP_INADMISSIBLE, and the published page must say so. A number produced by a compromised grader is not a small error to be discounted; it carries no information at all. Integer, deterministic. No hw writes (Rule 26). license_tier: ORIGINAL module: nishi-core.research.paper_panel depends: nx_referee.nx, nx_referee_v2.nx, nx_peer_review.nx genealogy_id: referee_v1_no_self_grade + referee_v2_r6_leakage + peer_review_dual_grade