Research: higher-arity dirty echoes use 152 calls at five-array cap (no new kappa)

research/space-capped-dirty-echo/CLAIM.json

@@ -0,0 +1,52 @@
+{
+  "track": "Recursive interfaces",
+  "assignment_seed": 58,
+  "randomness": "Python secrets.randbelow(100), operating-system CSPRNG, one draw",
+  "base_commit": "66849e1475c930fd7732b1a0067ffee7e1e95f60",
+  "started_utc": "2026-10-13T04:34:32Z",
+  "claimed_utc": "2026-10-13T04:39:00Z",
+  "expected_finish_utc": "2026-10-13T05:34:32Z",
+  "verification_reserved_from_utc": "2026-10-13T05:19:32Z",
+  "coordination": "GitHub pinnedIssues is empty; issue 650 is a discussion request with no comments. Claim retained locally as instructed.",
+  "initial_overlap": "528 open PRs through #708; reviewed titles and relevant bodies, review ledger, selected construction/review and prior negative-search index",
+  "candidates": [
+    "Defer owner-boundary residual restoration in the selected #626 allocation: close to #675 and existing whole-owner assumptions",
+    "Optimize arbitrary contiguous dirty-echo partitions and arities under a hard scratch cap: selected; #689 explicitly excludes unbalanced splits and unequal costs",
+    "Split width-9 residual hosting with a different boundary contract: close to #665's fully priced negative result"
+  ],
+  "object": "PR673 multiway dirty echo, with arbitrary contiguous factor partitions and arities under a fixed peak borrowed-array budget",
+  "method": "Exact interval dynamic programming, independent exhaustive small-tree oracle, nonlinear restoration and noncommutative operator checks; complete leaf/phase-access/precision recurrence",
+  "seed_use": "58 mod 8 = 2 selects radix 16 from [4,9,16,25,27,32,49,64]; use its 15 factors and #689's five-array budget as the primary comparison",
+  "deliverable": "One isolated research package with optimality scoped to sequential multiway echo trees, literal schedules, reproducible certificates, negative controls and no new kappa claim",
+  "stopping_rule": "One planner and its fixed finite verification domain; stop after exact comparison and verification. Do not pursue supplier integration or another mechanism.",
+  "dependency_heads": {
+    "628": "0768f97d06c623b8211e651543c66135796b62bf",
+    "673": "bfff94e78f0e75ebd2593e3cc7710ae5290e8be3",
+    "689": "0dedef12f05f9a06faf9d4e1c8b1462c0e3bff3a"
+  },
+  "contributor": "Koroffstrogov, with substantial OpenAI Codex assistance",
+  "status": "completed",
+  "result_update_utc": "2026-10-13T05:01:20Z",
+  "actual_verification_and_handoff_start_utc": "2026-10-13T04:52:00Z",
+  "result": "Exact all-arity, unbalanced contiguous echo-tree planner: 160 -> 152 complete calls at five arbitrary dirty arrays; weighted radix-16 phase accesses 3840 -> 3656. No new kappa.",
+  "result_link": "https://github.com/Koroffstrogov/integer-mult-bounds/tree/research/space-capped-dirty-echo/research/space-capped-dirty-echo",
+  "final_overlap": {
+    "open_prs": 544,
+    "through_pr": 725,
+    "pinned_issue_count": 0,
+    "closest_concurrent": {
+      "714": "efad8a40094ab453c0525efef0b0eff2a5d8b850: requires one clean root array, unlike the all-dirty pool here",
+      "689": "0dedef12f05f9a06faf9d4e1c8b1462c0e3bff3a: balanced ternary only; credited baseline",
+      "673": "156f259b99723468cfdd2b5d47d0ba894f09bea7: new bit-role sharing screen; all three consumed files unchanged from pinned bfff94e"
+    }
+  },
+  "publication_overlap_supplement": {
+    "through_pr": 745,
+    "inspected": {
+      "730": "d222e348179cc02a54d6d26dbd2838944f62c6e5",
+      "744": "576c6b20065fff3402c1eed2ad74c5c7d03029e8"
+    },
+    "distinct_result": "Higher arities already in the original claimed scope; pinned #730 binary/ternary optimizer gives 160 while all-arity gives 152 at n=15, cap=5. Actual weighted radix-16 replay adds 3840 -> 3656 phase reads. #744 independently checks the narrower ternary class. Both credited, no priority claim for unbalanced or weighted optimization.",
+    "reproduce_comparison": "python3 -I -O -B research/space-capped-dirty-echo/compare_concurrent.py"
+  }
+}

research/space-capped-dirty-echo/CONCURRENT-SOURCES.json

@@ -0,0 +1,16 @@
+{
+  "sources": [
+    {
+      "path": "vendor/pr730/optimize.py",
+      "commit": "d222e348179cc02a54d6d26dbd2838944f62c6e5",
+      "source_path": "research/ordered-dirty-echo/optimize.py",
+      "sha256": "d2cff4c43cfec569454c43af28b2a4fb1fe0b8521722c6c5003ba062d89f526c"
+    },
+    {
+      "path": "vendor/pr730/NOTICE.md",
+      "commit": "d222e348179cc02a54d6d26dbd2838944f62c6e5",
+      "source_path": "research/ordered-dirty-echo/NOTICE.md",
+      "sha256": "492220b4a4fa193615a04a475622c0bd2aca991a92ddcaccfba7f9b7750dddcc"
+    }
+  ]
+}

research/space-capped-dirty-echo/CONCURRENT.json

@@ -0,0 +1,54 @@
+{
+  "pr730_commit": "d222e348179cc02a54d6d26dbd2838944f62c6e5",
+  "status": "PASS_HIGHER_ARITY_COMPARISON",
+  "uniform": [
+    {
+      "all_arity_calls": 152,
+      "cap": 5,
+      "factors": 15,
+      "pr730_calls": 160
+    },
+    {
+      "all_arity_calls": 128,
+      "cap": 6,
+      "factors": 15,
+      "pr730_calls": 144
+    },
+    {
+      "all_arity_calls": 448,
+      "cap": 6,
+      "factors": 27,
+      "pr730_calls": 464
+    },
+    {
+      "all_arity_calls": 1600,
+      "cap": 8,
+      "factors": 58,
+      "pr730_calls": 1728
+    }
+  ],
+  "weighted": {
+    "all_arity_calls": 152,
+    "all_arity_phase_reads": 3656,
+    "cap": 5,
+    "pr730_calls": 160,
+    "pr730_phase_reads": 3840,
+    "weights": [
+      23,
+      28,
+      28,
+      28,
+      23,
+      23,
+      23,
+      23,
+      23,
+      23,
+      23,
+      23,
+      23,
+      28,
+      23
+    ]
+  }
+}