Tooling: 5.35× faster production leech_minor via base-256 byte lanes with identical receipts

certificates/affine-vector-screens.json

@@ -143,7 +143,7 @@
   "source_sha256": {
     "docs/research/affine-vector-screens.md": "22e5beab6df8127fc4377142768fc1a0a3317d7e8fba52c089670ecf1aa076b6",
     "scripts/audit_affine_vectors.py": "a3fccaeadb9a5db72c33787a055a54374504a6aa231907080396b2cf822e94a7",
-    "scripts/experiments/affine_vector_screens.py": "756778eeca27f490adc7354a122dc50c40375156259991469c47a90949118ea1",
+    "scripts/experiments/affine_vector_screens.py": "c499289fc25b5d180df22f70694a7b1c4cfed68f00a9a59bbf4f7ae1b52327a9",
     "scripts/experiments/affine_vector_witnesses.json": "672a4318c48f5286ce088d34736c38957b7f662a5f753d440dbc638e7da5f62a",
     "scripts/experiments/cube_core_family.py": "e58f1f4067bda2d4838343f42ec2505316f006561f0702e869847d88b85413a8",
     "scripts/experiments/single_intersection.py": "6874d41b75590369182ed61e0c6da5b547e035e0e43b50ea2b39e1f3861dca6d"

research/leech-minor-production/NOTICE.md

@@ -0,0 +1,20 @@
+# Attribution and license
+
+Prepared by Wentao Li ([VictorLiwentao](https://github.com/VictorLiwentao),
+[@Is__Victor](https://x.com/Is__Victor), wentaol4@illinois.edu) with substantial
+AI assistance for orientation, overlap search, profiling, production wiring,
+packaging, benchmarking and writing. Not independent human peer review.
+Apache-2.0; see the repository root `LICENSE`.
+
+The exact base-256 byte-lane algorithm and its no-carry argument were introduced
+by **Jerry Chen (@CubeJerry)** in the opt-in adapter
+`research/leech-minor-byte-lanes/` (incorporated
+[CrocSwap/integer-mult-bounds#898](https://github.com/CrocSwap/integer-mult-bounds/pull/898);
+see that package's NOTICE.md). This package wires an equivalent check into the
+production `scripts/experiments/affine_vector_screens.py:leech_minor` path used
+by `make verify-certificates`.
+
+The original affine/vector screens, Golay inventory, Leech line generators,
+rational span witnesses, binary elimination and nine-screen certificate are by
+**Douglas R. Colkitt**. Upstream material retains its own source notices and
+license.

research/leech-minor-production/README.md

@@ -0,0 +1,103 @@
+# Production byte-lane Leech-minor rows
+
+**Tooling improvement; no new kappa.** At base
+`0fdd7d82de03e43c9e5a1000dfbcd380d1b687d2`,
+`scripts/experiments/affine_vector_screens.py:leech_minor` still built each
+binary row with the historical pairwise sparse-dot genexpr. Jerry Chen's
+opt-in adapter `research/leech-minor-byte-lanes/` (incorporated
+[CrocSwap/integer-mult-bounds#898](https://github.com/CrocSwap/integer-mult-bounds/pull/898))
+already proved an exact base-256 byte-lane algorithm with identical receipts.
+This package wires that equivalent check into the **production** source used by
+`scripts/audit_affine_vectors.py` / `make verify-certificates`.
+
+Only the source pin for `scripts/experiments/affine_vector_screens.py` changes
+inside `certificates/affine-vector-screens.json`. Every matrix entry is still
+materialised; Golay inventory, norms, rational span witnesses, binary
+elimination, polynomial-code screens and the retained historical `2^-31`
+language are unchanged. Selected frontier untouched.
+
+## Reproduce
+
+Python **3.12**, standard library only, from the repository root:
+
+```sh
+python3.12 -I -B research/leech-minor-production/verify.py
+python3.12 -I -O -B research/leech-minor-production/verify.py
+python3.12 -B scripts/research_package.py --package research/leech-minor-production
+python3.12 -I -B research/leech-minor-production/benchmark.py --rounds 5
+```
+
+Expected:
+
+```text
+PASS_LEECH_MINOR_PRODUCTION leech_cases=5 code_cases=4 certs=1 no_new_kappa
+PASS_REGISTERED_RESEARCH_PACKAGE research/leech-minor-production outcome=completed
+PASS_IDENTICAL_LEECH_MINOR_PRODUCTION_RECEIPTS
+```
+
+Default verify is read-only. Use `--write-certificate` only after editing the
+production source. Timing is host-descriptive and is not a correctness gate.
+
+On Linux / CPython 3.12.3, five alternating fresh-process pairs gave median
+**leech-only 5.579 → 1.043 s (5.35×)** and **complete audit 9.465 → 4.213 s
+(2.25×)**. Samples are in [`RESULT.json`](RESULT.json).
+
+## Why the byte lanes are equivalent
+
+Coordinates of every Leech sample vector lie in `[-4,4]`. For each coordinate
+`k`, pack nonnegative base-256 integers
+
+```text
+P_k = sum_j (v_j[k] + 4) 256^j
+M_k = sum_j (4 - v_j[k]) 256^j
+```
+
+For a row vector `u` with `||u||_1 <= 31` (true of every historical Leech row:
+l1 norms 8, 16 or 26),
+
+```text
+S_u = sum_{u[k]>0} u[k] P_k + sum_{u[k]<0} (-u[k]) M_k
+```
+
+has byte `j` equal to `dot(u,v_j) + 4 ||u||_1`, with every digit at most
+`8*31 < 256` and no inter-column carries. Decoding bytes and comparing to
+`shift + 4 ||u||_1`, while forcing the diagonal, returns exactly
+`C[i,j] = (i == j or dot(v_i,v_j) == shift)` in the original column order.
+This is the same argument as `research/leech-minor-byte-lanes/README.md`.
+
+## Overlap and credit
+
+Closest priors:
+
+- [Jerry Chen / leech-minor-byte-lanes](../leech-minor-byte-lanes/) — opt-in
+  byte-lane adapter (#898 incorporated); this package is the production wiring.
+- [CrocSwap/integer-mult-bounds#756](https://github.com/CrocSwap/integer-mult-bounds/pull/756)
+  — identical-receipt tooling quality bar.
+- Open leave-alone tooling on other objects (#1922 joint frames, #2054/#2150
+  shared/mixed-point, #2100 ternary, #2117 DisjointTensor.frame_data, #2194
+  production verify_factor, #1861 parity-side) was not reopened.
+
+Douglas R. Colkitt authored the affine/vector screens and original checker.
+Jerry Chen authored the byte-lane algorithm. Prepared by Wentao Li
+([VictorLiwentao](https://github.com/VictorLiwentao),
+[@Is__Victor](https://x.com/Is__Victor)) with substantial AI assistance, under
+Apache-2.0.
+
+## Open next questions
+
+1. **Question:** Should the four polynomial-code screens in the same audit
+   receive a similar production packing pass?
+   **Why it matters:** After Leech minors shrink, the unchanged cubic-code path
+   is a larger share of the complete audit.
+   **First step:** Profile `audit_affine_vectors.certificate` with production
+   lanes already merged.
+   **Blocked by:** No open identical-receipt adapter yet for
+   `polynomial_code_witness`.
+
+2. **Question:** Should `research/leech-minor-byte-lanes/` note that production
+   now ships the same algorithm?
+   **Why it matters:** That package's source-admission string match expects the
+   historical pairwise block.
+   **First step:** Add a one-line status note without editing pinned evidence.
+   **Blocked by:** Maintainer preference on touching incorporated historical
+   packages.

research/leech-minor-production/RESULT.json

@@ -0,0 +1,142 @@
+{
+  "ai_assistance": "Substantial AI assistance for orientation, overlap search, profiling, implementation, packaging, benchmarking and writing. Not independent human peer review.",
+  "base_commit": "0fdd7d82de03e43c9e5a1000dfbcd380d1b687d2",
+  "complete_audit": {
+    "median_baseline_seconds": 9.46477570900015,
+    "median_optimized_seconds": 4.212500120000186,
+    "median_ratio": 2.246830964837986,
+    "rounds": 5,
+    "samples": [
+      {
+        "baseline_seconds": 9.46477570900015,
+        "optimized_seconds": 4.516600023000137,
+        "order": [
+          "baseline",
+          "optimized"
+        ],
+        "ratio": 2.0955532171992513,
+        "round": 0
+      },
+      {
+        "baseline_seconds": 12.871481419000247,
+        "optimized_seconds": 4.212500120000186,
+        "order": [
+          "optimized",
+          "baseline"
+        ],
+        "ratio": 3.0555444634621587,
+        "round": 1
+      },
+      {
+        "baseline_seconds": 10.286713764999604,
+        "optimized_seconds": 5.023797816999831,
+        "order": [
+          "baseline",
+          "optimized"
+        ],
+        "ratio": 2.047597084856958,
+        "round": 2
+      },
+      {
+        "baseline_seconds": 9.034710279999672,
+        "optimized_seconds": 4.105965091000144,
+        "order": [
+          "optimized",
+          "baseline"
+        ],
+        "ratio": 2.2003865302710035,
+        "round": 3
+      },
+      {
+        "baseline_seconds": 8.674803519999841,
+        "optimized_seconds": 3.795328679999784,
+        "order": [
+          "baseline",
+          "optimized"
+        ],
+        "ratio": 2.2856527725025213,
+        "round": 4
+      }
+    ],
+    "scope": "audit"
+  },
+  "engines": {
+    "baseline": "research/leech-minor-production/baseline.py",
+    "optimized": "scripts/experiments/affine_vector_screens.py"
+  },
+  "environment": {
+    "cpu_affinity_set": false,
+    "executable": "/usr/bin/python3.12",
+    "implementation": "CPython",
+    "os_caches_flushed": false,
+    "platform": "Linux-6.12.94+-x86_64-with-glibc2.39",
+    "python": "3.12.3 (main, Aug 31 2026, 10:18:26) [GCC 13.3.0]"
+  },
+  "leech_only": {
+    "median_baseline_seconds": 5.579358917999798,
+    "median_optimized_seconds": 1.0430829469996752,
+    "median_ratio": 5.348912024732329,
+    "rounds": 5,
+    "samples": [
+      {
+        "baseline_seconds": 5.555424422000215,
+        "optimized_seconds": 0.6638417809999737,
+        "order": [
+          "baseline",
+          "optimized"
+        ],
+        "ratio": 8.368597128116638,
+        "round": 0
+      },
+      {
+        "baseline_seconds": 5.502019425999606,
+        "optimized_seconds": 1.0430829469996752,
+        "order": [
+          "optimized",
+          "baseline"
+        ],
+        "ratio": 5.274766922252559,
+        "round": 1
+      },
+      {
+        "baseline_seconds": 5.914984036000078,
+        "optimized_seconds": 0.8409578649998366,
+        "order": [
+          "baseline",
+          "optimized"
+        ],
+        "ratio": 7.033627108061148,
+        "round": 2
+      },
+      {
+        "baseline_seconds": 5.984793322999849,
+        "optimized_seconds": 1.187062658999821,
+        "order": [
+          "optimized",
+          "baseline"
+        ],
+        "ratio": 5.041682743219666,
+        "round": 3
+      },
+      {
+        "baseline_seconds": 5.579358917999798,
+        "optimized_seconds": 1.1603133880003043,
+        "order": [
+          "baseline",
+          "optimized"
+        ],
+        "ratio": 4.808493098244192,
+        "round": 4
+      }
+    ],
+    "scope": "leech"
+  },
+  "no_new_kappa": true,
+  "predecessor": "research/leech-minor-byte-lanes/",
+  "source_sha256": {
+    "baseline.py": "756778eeca27f490adc7354a122dc50c40375156259991469c47a90949118ea1",
+    "research/leech-minor-byte-lanes/lanes.py": "2c242d4284e1c8520cbc0cfbbb20ddf186c662e5f6173bed400bbd4022df34c7",
+    "scripts/experiments/affine_vector_screens.py": "c499289fc25b5d180df22f70694a7b1c4cfed68f00a9a59bbf4f7ae1b52327a9"
+  },
+  "status": "PASS_IDENTICAL_LEECH_MINOR_PRODUCTION_RECEIPTS"
+}

research/leech-minor-production/SOURCES.json

@@ -0,0 +1,60 @@
+{
+  "schema_version": 1,
+  "base_commit": "0fdd7d82de03e43c9e5a1000dfbcd380d1b687d2",
+  "files": [
+    {
+      "path": "docs/research/agent-task-catalog.json",
+      "usage": "provenance",
+      "git_blob": "e8870e5917057d59b730eb14721a778639eca050",
+      "sha256": "d46eb6d5b3e279428753d770653663fbdf1063e55c255ea073d768f344927ef7"
+    },
+    {
+      "path": "certificates/selected-result.json",
+      "usage": "provenance",
+      "git_blob": "ae6a99f28aa468649eb66a361e764d769363bf11",
+      "sha256": "10b9e340fd8e354a37a0c4813c90e5de1dd42044b5f6aca4a3fc5a42b685fa32"
+    },
+    {
+      "path": "certificates/affine-vector-screens.json",
+      "usage": "provenance",
+      "git_blob": "7553cd304e62c7a3c9136046663d418cb068f0e8",
+      "sha256": "e7920c3305bd5e3058d2614fadc064ec1f658029183325f600c7ee4b56a7906e"
+    },
+    {
+      "path": "docs/research/affine-vector-screens.md",
+      "usage": "provenance",
+      "git_blob": "13f454478b554408cb55fdb31d514162b5c4b241",
+      "sha256": "22e5beab6df8127fc4377142768fc1a0a3317d7e8fba52c089670ecf1aa076b6"
+    },
+    {
+      "path": "scripts/audit_affine_vectors.py",
+      "usage": "provenance",
+      "git_blob": "9f5e3dff5e7c0444d160fec3a773fd702b0bc569",
+      "sha256": "a3fccaeadb9a5db72c33787a055a54374504a6aa231907080396b2cf822e94a7"
+    },
+    {
+      "path": "scripts/experiments/affine_vector_screens.py",
+      "usage": "provenance",
+      "git_blob": "0686b47d4844bedc9dddfacb0bff52ec552120a2",
+      "sha256": "756778eeca27f490adc7354a122dc50c40375156259991469c47a90949118ea1"
+    },
+    {
+      "path": "scripts/experiments/affine_vector_witnesses.json",
+      "usage": "input",
+      "git_blob": "0e9a488ac4912cbf2bb83894c1b6c9b607cc02cd",
+      "sha256": "672a4318c48f5286ce088d34736c38957b7f662a5f753d440dbc638e7da5f62a"
+    },
+    {
+      "path": "research/leech-minor-byte-lanes/lanes.py",
+      "usage": "input",
+      "git_blob": "095173fba90d619d8ff9cc46e28431af85fee65d",
+      "sha256": "2c242d4284e1c8520cbc0cfbbb20ddf186c662e5f6173bed400bbd4022df34c7"
+    },
+    {
+      "path": "research/leech-minor-byte-lanes/README.md",
+      "usage": "provenance",
+      "git_blob": "f6f50385e8f08aea1b1eda245681b4add515cdc2",
+      "sha256": "5ac99737af9ab3422faef2825d8e592d32d7dcb90fff83cd025702e895fead26"
+    }
+  ]
+}