Scoped negative result: CK target-forest floor holds for all dyadic weights (no new kappa)

research/ck-all-dyadic-forest-seed23/CLAIM.json

@@ -0,0 +1,25 @@
+{
+  "track": "complex supplier",
+  "assignment_seed": 23,
+  "seed_source": "secrets.randbelow(100), operating-system randomness",
+  "base_commit": "1b59b624d21924f8e3537cdc1ac52d3be6eda2fe",
+  "scope": "Historical PR626 CK compensation, arbitrary nonzero dyadic single-parent target forest weights, independent binary-Horner coefficient lowering",
+  "candidates": [
+    {"task": "Selected k8 rank-seven ends / owner packing", "decision": "Rejected: crowded by PR1090 and PR1117 through PR1169"},
+    {"task": "Contracted factory fixed-order diagonal scales", "decision": "Rejected: complex-factory-scale-frontier27 closes this class; PR1064/1106 also overlap"},
+    {"task": "CK nonunit target-parent coefficients", "decision": "Chosen after coefficient-domain refinement below"}
+  ],
+  "pivot": "Initial fixed-topology signed-power-of-two hypothesis had no gains, but deeper all-state PR-body search identified PR964 already closing all power-of-two CK forests. That duplicate was abandoned. Exactly one pivot changes the coefficient class to arbitrary dyadic numerators, explicitly left open by PR964. PR961 supplies credited universal tails for the different X4 matrix, not CK census results.",
+  "domain": "All 12432 ordered distinct CK target pairs and 1086 signed coefficients n/8, 1<=abs(n)<=543; written universal tails for reduced denominator exponent >=4 or magnitude >=68",
+  "stopping_rule": "Complete deterministic finite core and tail proof, no optimizer restarts or time-limited search",
+  "method": "Exact integer coefficient-pair histograms, fully paid root/unit dominance, inherited forest exchange lemma, independent Prim/cycle check, all-column rational endpoint controls",
+  "optimistic_budget": "E(i,j,lambda)=R(C_i-lambda*C_j)+2H(lambda). Both target-boundary shears, residual coefficient work and clears charged. Other local stages and all recursive children fixed; any exclusion cannot change kappa.",
+  "expected_deliverable": "Exact scoped exclusion or concrete nonunit improving edge with complete cost; no hosted supplier or kappa claim",
+  "coordination": "No pinned sprint coordination issue existed; open issues913 and650 were unrelated. Discussions1163,1166,1167 had no comments/claims on entry. Local claim and parent reservation; parent approved the overlap-driven pivot.",
+  "expected_finish": "2026-10-13T22:30:00Z",
+  "status": "complete; focused research PR submitted",
+  "result": "All-dyadic CK target-forest floor remains16364. Complete13501152-case census has minimum valid domination margin63;941 nonunit cases beat their root (minimum−116), requiring the same-parent exchange. All716 columns in both orientations,10 mathematical controls, full independent census oracle and3 packaging rejection controls pass.",
+  "pull_request": "https://github.com/CrocSwap/integer-mult-bounds/pull/1220",
+  "verification": "python3 -I -O -B research/ck-all-dyadic-forest-seed23/verify.py && python3 -I -O -B research/ck-all-dyadic-forest-seed23/independent.py; separate test_package.py",
+  "scope_limits": "Historical independent binary-Horner one-parent forest class only. No changed supplier, native/global admission, precision improvement or new kappa."
+}