Tooling: 15× faster production sparse-fusion screen via spectator orbits with identical receipts

research/sparse-fusion-orbits-production/NOTICE.md

@@ -0,0 +1,17 @@
+# 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 spectator-orbit reduction and its permutation-conjugation argument
+were introduced by **ToikY** in the opt-in adapter
+`research/sparse-fusion-orbits/` (see that package's NOTICE.md and PROOF.md).
+This package wires an equivalent check into the production
+`scripts/audit_sparse_fusion.py:central_gate_screen` path used by `make verify`.
+
+The original sparse/fusion finite screen, phase-group identities and scientific
+receipt are by **Douglas R. Colkitt**. Upstream material retains its own source
+notices and license.

research/sparse-fusion-orbits-production/README.md

@@ -0,0 +1,96 @@
+# Production spectator orbits for the sparse-fusion screen
+
+**Tooling improvement; no new kappa.** At base
+`0fdd7d82de03e43c9e5a1000dfbcd380d1b687d2`,
+`scripts/audit_sparse_fusion.py:central_gate_screen` still summed exact binary
+ranks over every weight-three indicator. ToikY's opt-in adapter
+`research/sparse-fusion-orbits/` already proved an exact spectator-orbit
+reduction with identical receipts. This package wires that equivalent check
+into the **production** source used by `make verify`.
+
+`certificates/sparse-fusion-audit.json` stays **byte-identical** (this audit has
+no `source_sha256` map). Every screen field, all 4,096 joint-frame inequalities,
+finite phase identities, candidate budgets and the retained historical language
+are unchanged. Selected frontier untouched.
+
+## Reproduce
+
+Python **3.12**, standard library only, from the repository root:
+
+```sh
+python3.12 -I -B research/sparse-fusion-orbits-production/verify.py
+python3.12 -I -O -B research/sparse-fusion-orbits-production/verify.py
+python3.12 -B scripts/research_package.py --package research/sparse-fusion-orbits-production
+python3.12 -I -B research/sparse-fusion-orbits-production/benchmark.py --rounds 5
+```
+
+Expected:
+
+```text
+PASS_SPARSE_FUSION_ORBITS_PRODUCTION screen_heights=[22, 25, 28] screen_cases=3 certs=1 no_new_kappa
+PASS_REGISTERED_RESEARCH_PACKAGE research/sparse-fusion-orbits-production outcome=completed
+PASS_IDENTICAL_SPARSE_FUSION_ORBITS_PRODUCTION_RECEIPTS
+```
+
+Default verify is read-only. Use `--write-certificate` only after editing the
+production source if regeneration is ever needed; the mathematical certificate
+must still match the base bytes. Timing is host-descriptive and is not a
+correctness gate.
+
+On Linux / CPython 3.12.3, five alternating fresh-process pairs gave median
+**screen-only 0.661 → 0.044 s (15.05×)** and **complete certificate 0.608 →
+0.060 s (10.07×)**. Retained samples are in [`RESULT.json`](RESULT.json).
+
+## Why the orbits are equivalent
+
+For each of the 64 symmetric three-coordinate blocks `S`, the screen builds
+`Q = diag(S, I_(h-3))`. Permutations fixing coordinates 0,1,2 satisfy
+`P Q P^T = Q`, so
+
+```text
+rank(Q + t t^T) = rank(Q + (P t)(P t)^T).
+```
+
+Weight-three indicators with the same low three bits form one orbit of size
+`C(h-3, 3-wt(low))`. Summing the eight weighted representatives therefore equals
+the historical sum over all `C(h,3)` triples. At `h=25`, unary ranks fall from
+147,200 to 512; the 128 frame ranks and 4,096 joint-pair ranks stay unchanged.
+This is the same argument as `research/sparse-fusion-orbits/PROOF.md`.
+
+## Overlap and credit
+
+Closest priors:
+
+- [ToikY / sparse-fusion-orbits](../sparse-fusion-orbits/) — opt-in orbit
+  adapter; 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, #2210 production leech_minor, #1861 parity-side)
+  was not reopened. Fresh overlap refresh over 464 open PRs found no competing
+  production rewrite of `central_gate_screen`.
+
+Douglas R. Colkitt authored the sparse/fusion screen and original checker.
+ToikY authored the orbit 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 `research/sparse-fusion-orbits/` note that production
+   now ships the same orbit reduction?
+   **Why it matters:** That package's source-admission pins expect the
+   historical pairwise enumeration block.
+   **First step:** Add a one-line status note without editing pinned evidence.
+   **Blocked by:** Maintainer preference on touching incorporated historical
+   packages.
+
+2. **Question:** After orbit wiring, is the remaining joint-pair double loop
+   the next productive production pass?
+   **Why it matters:** With unary ranks reduced, joint pairs dominate residual
+   screen time.
+   **First step:** Profile production `central_gate_screen` with orbits merged.
+   **Blocked by:** No open identical-receipt adapter yet for the joint-pair
+   stage.

research/sparse-fusion-orbits-production/RESULT.json

@@ -0,0 +1,142 @@
+{
+  "ai_assistance": "Substantial AI assistance for orientation, overlap search, profiling, production wiring, packaging, benchmarking and writing. Not independent human peer review.",
+  "base_commit": "0fdd7d82de03e43c9e5a1000dfbcd380d1b687d2",
+  "complete_certificate": {
+    "median_baseline_seconds": 0.6076670360000662,
+    "median_optimized_seconds": 0.060372408999910476,
+    "median_ratio": 10.065310397022046,
+    "rounds": 5,
+    "samples": [
+      {
+        "baseline_seconds": 0.6030950699999948,
+        "optimized_seconds": 0.0608979090000048,
+        "order": [
+          "baseline",
+          "optimized"
+        ],
+        "ratio": 9.903378948527564,
+        "round": 0
+      },
+      {
+        "baseline_seconds": 0.7542092740000044,
+        "optimized_seconds": 0.05746237099992868,
+        "order": [
+          "optimized",
+          "baseline"
+        ],
+        "ratio": 13.125272432648845,
+        "round": 1
+      },
+      {
+        "baseline_seconds": 0.6360726619999468,
+        "optimized_seconds": 0.060372408999910476,
+        "order": [
+          "baseline",
+          "optimized"
+        ],
+        "ratio": 10.535817147877767,
+        "round": 2
+      },
+      {
+        "baseline_seconds": 0.6076670360000662,
+        "optimized_seconds": 0.060805378999930326,
+        "order": [
+          "optimized",
+          "baseline"
+        ],
+        "ratio": 9.993639477204187,
+        "round": 3
+      },
+      {
+        "baseline_seconds": 0.581783529000063,
+        "optimized_seconds": 0.05653849500004071,
+        "order": [
+          "baseline",
+          "optimized"
+        ],
+        "ratio": 10.290042722213318,
+        "round": 4
+      }
+    ],
+    "scope": "certificate"
+  },
+  "engines": {
+    "baseline": "research/sparse-fusion-orbits-production/baseline.py",
+    "optimized": "scripts/audit_sparse_fusion.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]"
+  },
+  "no_new_kappa": true,
+  "predecessor": "research/sparse-fusion-orbits/",
+  "screen_only": {
+    "median_baseline_seconds": 0.6611210399998981,
+    "median_optimized_seconds": 0.04393979999997555,
+    "median_ratio": 15.046063932932467,
+    "rounds": 5,
+    "samples": [
+      {
+        "baseline_seconds": 0.61670235400004,
+        "optimized_seconds": 0.04277718100001948,
+        "order": [
+          "baseline",
+          "optimized"
+        ],
+        "ratio": 14.416619786136893,
+        "round": 0
+      },
+      {
+        "baseline_seconds": 0.672001172000023,
+        "optimized_seconds": 0.0456580989999793,
+        "order": [
+          "optimized",
+          "baseline"
+        ],
+        "ratio": 14.718115443227884,
+        "round": 1
+      },
+      {
+        "baseline_seconds": 0.6611210399998981,
+        "optimized_seconds": 0.04393979999997555,
+        "order": [
+          "baseline",
+          "optimized"
+        ],
+        "ratio": 15.046063932932467,
+        "round": 2
+      },
+      {
+        "baseline_seconds": 0.6881904790000135,
+        "optimized_seconds": 0.04753297199999906,
+        "order": [
+          "optimized",
+          "baseline"
+        ],
+        "ratio": 14.478170626486959,
+        "round": 3
+      },
+      {
+        "baseline_seconds": 0.6025585070000261,
+        "optimized_seconds": 0.04304783800000678,
+        "order": [
+          "baseline",
+          "optimized"
+        ],
+        "ratio": 13.997416246547184,
+        "round": 4
+      }
+    ],
+    "scope": "screen"
+  },
+  "source_sha256": {
+    "baseline.py": "bbaa32f9786f083409e363b081bfb7a1e597c8d32e4183c13a4105c2170934ce",
+    "research/sparse-fusion-orbits/accelerator.py": "17e5c373f6a45afb47e99219edec7f3aaf76facaeb0c42797683a0690d1c0fcd",
+    "scripts/audit_sparse_fusion.py": "6d1997fba802a8fc41099996b29e3add5d2b4f904da02548a98bed32c50aa773"
+  },
+  "status": "PASS_IDENTICAL_SPARSE_FUSION_ORBITS_PRODUCTION_RECEIPTS"
+}

research/sparse-fusion-orbits-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/sparse-fusion-audit.json",
+      "usage": "input",
+      "git_blob": "584ea8d56bbc3c6c12867f9551b425764fbe653d",
+      "sha256": "b7f0976e757998011f16022181246ad1b70db5b17ed32167f7e92368d2b43ee6"
+    },
+    {
+      "path": "docs/research/sparse-fusion-audit.md",
+      "usage": "provenance",
+      "git_blob": "32a3553a9e8a5072f1b669130e628d92ca868b5c",
+      "sha256": "631c852e068adf8d2943bf8f7373d822f423c2180ed0c8d413188a30ff51d75a"
+    },
+    {
+      "path": "scripts/audit_sparse_fusion.py",
+      "usage": "provenance",
+      "git_blob": "4e8e789d686b85a49fc12d36f87619eed705d6a8",
+      "sha256": "bbaa32f9786f083409e363b081bfb7a1e597c8d32e4183c13a4105c2170934ce"
+    },
+    {
+      "path": "tests/test_sparse_fusion.py",
+      "usage": "provenance",
+      "git_blob": "6c0fa720b4c67db5cee64dba962ec920ea21852c",
+      "sha256": "ad078502dbbcff04147665aed61609b31dc09a8e9a404d4e344aa5d462f568eb"
+    },
+    {
+      "path": "research/sparse-fusion-orbits/accelerator.py",
+      "usage": "input",
+      "git_blob": "24531ac40bdf283cab9d82f2a46d761ca9f03911",
+      "sha256": "17e5c373f6a45afb47e99219edec7f3aaf76facaeb0c42797683a0690d1c0fcd"
+    },
+    {
+      "path": "research/sparse-fusion-orbits/PROOF.md",
+      "usage": "provenance",
+      "git_blob": "118374b851be7164b05d476fdb0e329eb1e54cc3",
+      "sha256": "9c4dab77ef666c0c5dd6fd868e539597b88587d25574dde86b61736b77b617a2"
+    },
+    {
+      "path": "research/sparse-fusion-orbits/README.md",
+      "usage": "provenance",
+      "git_blob": "890e47e856f6d969843a2d8be57aaa5f98f281ac",
+      "sha256": "e25eb7c087e6b80dd6a191677720818dbdd65a01331b0e557d901c115ed24447"
+    }
+  ]
+}

research/sparse-fusion-orbits-production/TASK.json

@@ -0,0 +1,70 @@
+{
+  "schema_version": 1,
+  "task_id": "replay-runtime",
+  "priority_revision": 4,
+  "base_commit": "0fdd7d82de03e43c9e5a1000dfbcd380d1b687d2",
+  "chooser_seed": "daa8525f8946aff290000e81ce05cf4b",
+  "original_card": {
+    "id": "replay-runtime",
+    "question": "Can one measured replay bottleneck be sped up without dropping coverage?",
+    "starting_points": [
+      "scripts/run_isolated_tests.py",
+      "scripts/verify_replay_coverage.py",
+      "certificates/selected-result.json"
+    ],
+    "methods": [
+      "Profile one isolated replay and optimize the measured bottleneck.",
+      "Design source-bound reuse with explicit dependency invalidation and negative controls."
+    ],
+    "acceptance_gate": "Report before/after wall time on the same fixture, retained checks and invalidation behavior.",
+    "first_step": {
+      "command": [
+        "python3",
+        "-B",
+        "scripts/research_baseline.py"
+      ],
+      "expected": "PASS_RESEARCH_BASELINE_BINDINGS_ONLY",
+      "inspect": "Read the selected reproduction command and choose a single replay stage for profiling.",
+      "scope": "Positive control / source-bound orientation only; this is not a new construction or full cold replay."
+    },
+    "first_experiment": "Measure one unchanged fixture, preserve check coverage, and test one optimization with invalidation controls.",
+    "bottleneck_link": "Faster trustworthy replays increase review throughput and let mechanism candidates be checked sooner.",
+    "allowed_variations": [
+      "Profile one stage or design one source-bound cache boundary."
+    ],
+    "scope_change_triggers": [
+      "Skipping coverage or comparing different fixtures to claim speedup."
+    ],
+    "overlap_terms": [
+      "replay runtime",
+      "cache invalidation",
+      "isolated tests"
+    ]
+  },
+  "scope_status": "refined",
+  "actual_scope": "Can the production sparse-fusion central_gate_screen (scripts/audit_sparse_fusion.py, exercised by make verify / certificates/sparse-fusion-audit.json) adopt the exact spectator-orbit weighted representatives from research/sparse-fusion-orbits/ so the shipped verify path is faster while every screen field, joint-pair inequality and complete certificate stays identical?",
+  "scope_change_reason": "The catalogue card asks for one measured replay bottleneck with before/after wall time and retained checks. Expertise override selected verifier-engineering after profiling make verify poles. Leave-alone objects (joint frames, shared-point, mixed-point verify, ternary frames, parity-side, DisjointTensor.frame_data, production verify_factor, leech_minor production, quadratic-affine) and deferred micro-opts were not reopened. ToikY's incorporated sparse-fusion-orbits adapter already proved an opt-in identical-receipt accelerator; production still enumerated all C(h,3) triples pairwise.",
+  "bottleneck_link": "Side result on review throughput: sparse-fusion is part of make verify. Faster identical central_gate_screen ranks do not change kappa or the binding complex supplier.",
+  "ai_assistance": "Substantial AI assistance for orientation, overlap search, profiling, production wiring, packaging, benchmarking and writing",
+  "outcome": "completed",
+  "evidence": "PASS_SPARSE_FUSION_ORBITS_PRODUCTION via research/sparse-fusion-orbits-production/verify.py (also under -O); PASS_IDENTICAL_SPARSE_FUSION_ORBITS_PRODUCTION_RECEIPTS via benchmark.py; certificates/sparse-fusion-audit.json remains byte-identical to the base commit.",
+  "author": "Wentao Li",
+  "email": "wentaol4@illinois.edu",
+  "github": "VictorLiwentao",
+  "budget_minutes": 60,
+  "stop_rule": "Ship an equivalence-preserving production central_gate_screen change with frozen baseline, identical screen and complete-certificate payloads, rejection/temp-corruption controls, unchanged certificate bytes, registered replay and a measured before/after median. Do not invent kappa or reopen leave-alone objects.",
+  "open_questions": [
+    {
+      "question": "Should research/sparse-fusion-orbits/ note that production now ships the same orbit reduction?",
+      "why_it_matters": "That opt-in package's source-admission pins expect the historical pairwise enumeration block.",
+      "first_step": "Add a one-line status note to that package README without editing its pinned evidence.",
+      "blocked_by": "Maintainer preference on touching incorporated historical packages."
+    },
+    {
+      "question": "After orbit wiring, is the remaining joint-pair double loop (4096 ranks) the next productive production pass?",
+      "why_it_matters": "With unary ranks reduced from 147200 to 512 at h=25, joint pairs dominate residual screen time.",
+      "first_step": "Profile production central_gate_screen with orbits merged and name any further exact symmetry.",
+      "blocked_by": "No open identical-receipt adapter yet for the joint-pair stage."
+    }
+  ]
+}