Independently derive all 3,552 G20 MAIN-bank target ranks

research/g20-main-target-ranks/README.md

@@ -0,0 +1,105 @@
+# G20 MAIN-bank target ranks
+
+**Independent finite source-bound audit; no new bound or supplier admission.**
+This answers open question 2 of Matarener's [PR5955](https://github.com/CrocSwap/integer-mult-bounds/pull/5955),
+head `52115896549090117a3f3101b17106cad4466408`. Every one of the 3,552 MAIN
+owners in each PR5679/PR5708 inventory agrees with a rank derived from its
+retained controller. The MAIN rank mass is 154,290 in each candidate.
+
+| Address-projector rank | MAIN targets per candidate |
+| --- | ---: |
+| 15, 22, 29, 36, 43, 50, 57 | 444 each |
+| 64, 127 | 222 each |
+
+## Derivation and its contract
+
+The retained source proof describes a central line and eighteen disjoint
+seven-dimensional stage spaces. The ambient address space is their direct
+sum, of dimension 127. Let z reflect the central line and let x_s reflect
+stage s's block, acting identically on every other block. These commuting
+involutions are the stated controller model, not newly admitted primitives.
+
+The checker first reconstructs all 32 labeled output polynomials from the
+154 physical events and both boundary programs. Coefficients are in F2;
+each term retains its original input-bank label. ADD is symmetric difference,
+SWAP is three ADDs, and multiplying an involution toggles its monomial bit.
+A LOCAL event applies the central shear and its stage factor; IDLE applies
+its listed stage factors. Every off-diagonal coefficient must cancel, and
+each remaining diagonal monomial must equal the source declaration. No
+predecessor Python module or rank routine is imported.
+
+For a surviving monomial z times the product of x_s over S, evaluate its
+sign on each of the 127 coordinate basis columns. It is negative exactly
+on the central line and the disjoint blocks indexed by S. Thus
+P = (I - R)/2 is diagonal, with an identity minor on those columns and all
+other columns zero. Its rank is exactly 1 + 7|S|. The code evaluates the
+signed columns and cross-checks this formula rather than trusting a target
+rank or the controller's invoice. This rank is the **address projector's
+native job rank**, not the rank of a full FUNCTION permutation matrix.
+
+Every bank has 111 ports. For each MAIN descriptor, the checker verifies
+owner = 111*bank + port, uniqueness, complete owner coverage, its individual
+rank, both candidate invoice histograms and the controller's MAIN histogram.
+Both retained controllers are byte-identical. The full per-bank list and
+15 rejecting controls are in [RESULT.json](RESULT.json).
+
+## Scope, evidence and reproduction
+
+Base: `e21a287a7c4883c24cec069113ac7cfd5efcb3ca`, priority revision 8.
+PR5679 source head: `715dda75fc4cefb1e492349fa0e331d5f15d0f53`.
+PR5708 source head: `829d01fc031f7323ffcfa423fdfd1f24dcbe3190`.
+Exact retained source paths, Git blobs and SHA256 pins are in SOURCES.json.
+The verifier authenticates every live input before parsing it. TASK.json
+records the explicit refinement from the current BIT card.
+
+```sh
+python3.12 -B scripts/research_package.py --package research/g20-main-target-ranks
+```
+
+Expected marker: `PASS_G20_MAIN_TARGET_RANKS`. Each registered command has
+a 60-second timeout and uses only the standard library. Normal and optimized
+commands recompute the result and reject changed diagonals, missing or duplicate
+stages, wrong ambient dimensions, changed boundary gates, unknown opcodes,
+wrong target ranks/ports/owners, operator labels or route obligations, an equal-rank bank alias, missing targets,
+a count-and-rank-mass-preserving invoice change, and changed source bytes.
+`--write-result` explicitly regenerates RESULT.json; ordinary replay writes
+no package artifact. Semantic controls operate on copies after authentication.
+
+This closes only the source-relative MAIN rank input, including the formal
+controller endpoint that PR5955 did not replicate. It does not reprove the
+physical orthogonal-block realization, native all-column/whole-job adapters,
+FRAMES geometry, OWNED erasure, paid query ranks, global stream, finite fees,
+compiler, precision/tape, modulus domain, ordinary-to-admitted BIT conversion
+or outer assembly. The five framework assumptions remain. No selected record,
+frozen predecessor or original invoice is changed. Independent human and
+formal proof-kernel review are unclaimed. Hosted CI is separate from local checks.
+
+## Open next questions
+
+Bind one actual MAIN bank's physical projector and whole-job adapter to its
+all-column, arbitrary-spectator contract. This matters for physical admission;
+start from the exact coordinate action above and its retained native wrapper.
+The inherited native/compiler/precision obligations require a separate audit.
+The structured question and blocker are in TASK.json.
+
+## Credit and overlap
+
+Matarener supplied the independent review and explicit open question. Ali
+Romman, Dugongue, ai-nanachi and the retained G20 controller/native lineage
+supplied the constructions. JoyboyBrian's earlier bit-paid-owner-binding
+already emits productive ranks from the supplied controller diagonal; this
+package adds a separate complete polynomial reconstruction and current
+per-owner comparison, rather than claiming the rank formula is new.
+DreamingOfClouds, eumemic, Cinna Davis and the retained compiler/ownership
+authors keep their original attribution; no predecessor source is copied.
+
+Remote overlap refresh covered 315 open PRs, 5,221 closed PRs and 5,694
+Discussion records. Sparse local notes covered 7/8,346 paths; exact-gap
+searches of pinned Git notes and relevant older owner-binding code supplement
+that limited coverage. No competing exact-gap derivation was identified;
+this is not a global novelty guarantee. Coordination:
+https://github.com/CrocSwap/integer-mult-bounds/discussions/1166#discussioncomment-19231258
+
+New checker and exposition by Dennis / antunishdPursuit with substantial
+OpenAI Codex assistance, no subagents. Apache-2.0. Orthogonal direct sums,
+reflection projectors and rank by an identity minor are standard methods.

research/g20-main-target-ranks/RESULT.json

@@ -0,0 +1,70 @@
+{
+  "banks": 32,
+  "ports_per_bank": 111,
+  "MAIN_targets_per_candidate": 3552,
+  "controller_physical_events": 154,
+  "ambient_rank": 127,
+  "bank_ranks": [
+    15,
+    15,
+    22,
+    22,
+    29,
+    29,
+    36,
+    36,
+    43,
+    43,
+    50,
+    50,
+    57,
+    57,
+    15,
+    15,
+    22,
+    22,
+    29,
+    29,
+    36,
+    36,
+    43,
+    43,
+    50,
+    50,
+    57,
+    57,
+    64,
+    64,
+    127,
+    127
+  ],
+  "MAIN_histogram": {
+    "15": 444,
+    "22": 444,
+    "29": 444,
+    "36": 444,
+    "43": 444,
+    "50": 444,
+    "57": 444,
+    "64": 222,
+    "127": 222
+  },
+  "rank_mass_per_candidate": 154290,
+  "rejected_controls": [
+    "declared diagonal changed",
+    "missing thick call",
+    "duplicate thick stage",
+    "wrong ambient dimension",
+    "wrong boundary gate",
+    "unknown physical opcode",
+    "wrong target rank",
+    "wrong MAIN operator",
+    "missing route obligation",
+    "duplicate original owner",
+    "wrong target port",
+    "equal-rank bank alias",
+    "missing MAIN target",
+    "count-and-mass-preserving wrong invoice",
+    "changed source bytes"
+  ]
+}

research/g20-main-target-ranks/SOURCES.json

@@ -0,0 +1,66 @@
+{
+  "schema_version": 1,
+  "base_commit": "e21a287a7c4883c24cec069113ac7cfd5efcb3ca",
+  "files": [
+    {
+      "path": "certificates/selected-result.json",
+      "usage": "provenance",
+      "git_blob": "581f8d87008354eb5b869f7ed9de1317ad837ce8",
+      "sha256": "2a0220e866417774606e0f991e41b9c1d26b1f43b97857238cf478dbc7d330b5"
+    },
+    {
+      "path": "docs/research/agent-task-catalog.json",
+      "usage": "provenance",
+      "git_blob": "3404fbdd5e7529a0674eef26114b5071e4586cec",
+      "sha256": "ead6cdc9d0af6d58e8a49744d8ec8c7c15e13d2ba93c276c434bcf9b51b93a81"
+    },
+    {
+      "path": "research/g20-conjugation-admission/INVOICE.json",
+      "usage": "input",
+      "git_blob": "08ed5c4b9b0b03e16cd74b35880839c08a7125f4",
+      "sha256": "a2a9559bdc3b3293eae111c927e02e7cccecbaa0d35299c352421a58b3609296"
+    },
+    {
+      "path": "research/g20-conjugation-admission/TARGET-OWNERS.json",
+      "usage": "input",
+      "git_blob": "5721485cc27c49600d5c4d8e54c815175403f5de",
+      "sha256": "2918dc7e7e36551be7c9b5ea91a4df516d29573b8fd35e5a290a8d9944812754"
+    },
+    {
+      "path": "research/g20-conjugation-admission/prior/prior/prior/source/PROOF.md",
+      "usage": "input",
+      "git_blob": "951ce90c64428d596c678cca25a072930c45b794",
+      "sha256": "ebee57b34141253a90b0a72d959546362f236b5ee03c94fc0ff2b99f58149d37"
+    },
+    {
+      "path": "research/g20-conjugation-admission/prior/prior/prior/source/vendor/g20/inputs/CONTROLLER.json",
+      "usage": "input",
+      "git_blob": "89c400d5970024b3debe870c36ce43889beed001",
+      "sha256": "39ac84fb848b908f7ba051fae4556d507900c421dc5d23501a9c0cb1d61fd08d"
+    },
+    {
+      "path": "research/g20-helper-split/INVOICE.json",
+      "usage": "input",
+      "git_blob": "a4d9214abaab86dcd597b9533549ae1fe85672b5",
+      "sha256": "7213965618c196f2eb5528d6fdee8deab9152beca301e7a3ecddd7f8847e4803"
+    },
+    {
+      "path": "research/g20-helper-split/TARGET-OWNERS.json",
+      "usage": "input",
+      "git_blob": "5721485cc27c49600d5c4d8e54c815175403f5de",
+      "sha256": "2918dc7e7e36551be7c9b5ea91a4df516d29573b8fd35e5a290a8d9944812754"
+    },
+    {
+      "path": "research/g20-helper-split/prior/prior/source/vendor/g20/inputs/CONTROLLER.json",
+      "usage": "input",
+      "git_blob": "89c400d5970024b3debe870c36ce43889beed001",
+      "sha256": "39ac84fb848b908f7ba051fae4556d507900c421dc5d23501a9c0cb1d61fd08d"
+    },
+    {
+      "path": "research/g20-independent-review/README.md",
+      "usage": "input",
+      "git_blob": "92deccb3ac4712b2047922fa9c426e046ebc7119",
+      "sha256": "d0a007f675f433372679f71686283a2b0150e9200cfe86a0d9b92d945489d41e"
+    }
+  ]
+}

research/g20-main-target-ranks/TASK.json

@@ -0,0 +1,77 @@
+{
+  "schema_version": 1,
+  "task_id": "bit-helper-contract",
+  "priority_revision": 8,
+  "base_commit": "e21a287a7c4883c24cec069113ac7cfd5efcb3ca",
+  "original_card": {
+    "id": "bit-helper-contract",
+    "question": "Can a changed G20 helper path or independent-target split improve the complete BIT moment without losing its physical endpoint?",
+    "starting_points": [
+      "docs/research/current-priorities.md",
+      "docs/research/research-family-status.md",
+      "certificates/selected-result.json",
+      "docs/research/pr-intake-evidence/pr5599-completion/PROOF-REVIEW.md",
+      "research/g20-helper-split/README.md",
+      "research/g20-conjugation-admission/README.md"
+    ],
+    "methods": [
+      "Construct and independently check one source-bound instance of the stated first experiment.",
+      "Seek an exact counterexample or target-specific complete-budget obstruction to that same proposed change."
+    ],
+    "acceptance_gate": "A full boundary contract and paid profile; unchanged scalar counts alone are insufficient. A scoped local result may be retained without a complete supplier. Expanding it into further construction rounds requires the family-level budget and obligation gate in research-family-status.md; an optimistic passing relaxation alone is not acceptance.",
+    "first_step": {
+      "command": [
+        "python3",
+        "-B",
+        "scripts/research_selected_baseline.py"
+      ],
+      "expected": "PASS_RESEARCH_BASELINE_BINDINGS_ONLY",
+      "inspect": "Read revision 8, the selected review and live exact-source handoffs. Then follow this card's bounded first_experiment.",
+      "scope": "Source-bound orientation only, not a supplier replay or theorem proof."
+    },
+    "first_experiment": "First read #5679/#5708 and their audits as pending component proposals. Choose one uncovered helper group/cut on an exact word; compare arbitrary-input native action, targets, owned storage, both orientations and full signed bill. Stop if the gain is only a scalar count or repaid by lost targets. Do not rerun the frozen #5609 common-frame domain without changing its hypothesis.",
+    "bottleneck_link": "Judge the actual current supplier contract and complete signed target/paid bill; frozen-profile ceilings require an applicability proof.",
+    "allowed_variations": [
+      "One exact source head, helper/cut region or quantified proposition and changed hypothesis; check live handoffs first. Different seeds are not different scopes."
+    ],
+    "scope_change_triggers": [
+      "Changing source head, helper/cut domain, operation alphabet or caller boundary without recording the new scope.",
+      "Treating an archived proposal or local numerical pass as complete supplier admission."
+    ],
+    "overlap_terms": [
+      "G20",
+      "helper split",
+      "5679",
+      "5708",
+      "independent target"
+    ],
+    "existing_checks": [
+      {
+        "scope": "Selected PR5599: complete helper-subspace construction and independent caller replay; five framework assumptions remain.",
+        "evidence": "docs/research/pr-intake-evidence/pr5599-completion/PROOF-REVIEW.md",
+        "remaining": "The next experiment must identify uncovered source-specific work. Successor archive intake does not transfer this admission.",
+        "status": "reviewed",
+        "source_commit": "09dd0965711ca7eeb169057a420abab1c9052137",
+        "selected_record_sha256": "2a0220e866417774606e0f991e41b9c1d26b1f43b97857238cf478dbc7d330b5"
+      }
+    ]
+  },
+  "scope_status": "refined",
+  "actual_scope": "Independently derive all 3,552 MAIN-bank target ranks in each exact PR5679/PR5708 inventory from the retained controller, addressing PR5955 open question 2.",
+  "scope_change_reason": "Explicit user assignment follows the largest unchecked input in PR5955; this is a target-rank audit, not a new helper search or supplier admission.",
+  "bottleneck_link": "Removes reliance on declared MAIN ranks in the independent G20 review by rebuilding the formal endpoint and its retained orthogonal-block address projector.",
+  "budget_minutes": 1200,
+  "stop_rule": "Stop at any polynomial, owner, rank or invoice discrepancy, or after every MAIN target in both frozen candidates and the rejection controls pass. Do not expand into native/compiler admission.",
+  "outcome": "completed",
+  "evidence": "verify.py, RESULT.json and the proof and scope in README.md.",
+  "author": "Dennis / antunishdPursuit, with substantial OpenAI Codex assistance",
+  "ai_assistance": "Substantial OpenAI Codex assistance for source inspection, derivation, implementation, tests and writing. No subagents or independent human review claimed.",
+  "open_questions": [
+    {
+      "question": "Can the physical MAIN projector and whole-job adapter contract be independently admitted for arbitrary columns and spectators?",
+      "why_it_matters": "The coordinate rank derivation assumes the retained direct-sum physical interpretation; it does not implement its ROW or FUNCTION adapters.",
+      "first_step": "Choose one actual MAIN bank and bind its coordinate projector to the retained native all-column/address-to-FUNCTION contract, preserving routes and spectators.",
+      "blocked_by": "Inherited native/compiler/precision contracts and the need for a separately scoped physical audit."
+    }
+  ]
+}

research/g20-main-target-ranks/VERIFY.json

@@ -0,0 +1,43 @@
+{
+  "schema_version": 1,
+  "status": "ready",
+  "python": "3.12",
+  "requirements": [],
+  "inputs": [
+    "research/g20-conjugation-admission/INVOICE.json",
+    "research/g20-conjugation-admission/TARGET-OWNERS.json",
+    "research/g20-conjugation-admission/prior/prior/prior/source/PROOF.md",
+    "research/g20-conjugation-admission/prior/prior/prior/source/vendor/g20/inputs/CONTROLLER.json",
+    "research/g20-helper-split/INVOICE.json",
+    "research/g20-helper-split/TARGET-OWNERS.json",
+    "research/g20-helper-split/prior/prior/source/vendor/g20/inputs/CONTROLLER.json",
+    "research/g20-independent-review/README.md"
+  ],
+  "commands": [
+    {
+      "argv": [
+        "{python}",
+        "-W",
+        "error",
+        "-I",
+        "-B",
+        "{package}/verify.py"
+      ],
+      "timeout_seconds": 60,
+      "checks": "Authenticate eight source inputs; independently replay all controller coefficients; derive 127-coordinate projector ranks; compare all 3,552 MAIN owners in both candidates; reject 15 semantic/source mutations."
+    },
+    {
+      "argv": [
+        "{python}",
+        "-W",
+        "error",
+        "-I",
+        "-O",
+        "-B",
+        "{package}/verify.py"
+      ],
+      "timeout_seconds": 60,
+      "checks": "Same exact derivation and rejecting controls with optimized Python; all invariants use explicit exceptions."
+    }
+  ]
+}