Skip to content
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
146 changes: 146 additions & 0 deletions research/frontier1094-residual-wrappers-seed62/README.md
Original file line number Diff line number Diff line change
@@ -0,0 +1,146 @@
# The selected frontier1094 word's 36 residual wrappers are exact

**Supplier-boundary audit: an independent, byte-bound check. No new κ, and no existing check is changed.**

- Track: 60–74 (adversarial audit), assignment seed 62, drawn with the OS CSPRNG via Python `secrets`.
- Card: `supplier-boundary-audit`, with the scope **refined** in [TASK.json](TASK.json).
- Base: `179089e4`. The selected record frontier1094 is unchanged.

## The obligation and where its acceptance came from

Lemma A's exact part (`research/e8-lemma-a-lean`, `research/e8-lemma-g-lean/PROOF.md` §4) concerns every helper that enters at a nonzero frame E or ends below FULL at a hyperplane U. Each such helper needs a stored wrapper (L, R) with

`L ∘ (C_[w] ⊗ I) ∘ R = T_end ∘ T_E⁻¹`, w = dim end − dim E,

where end is FULL, or U for a hyperplane end. For the selected gauge44 provider these are the 36 records in `RESIDUAL-ADAPTERS.json`, raw sha256 `383f40aa`.

The selected chain binds the bytes of that file, but not their exactness:
- **The bytes are pinned.** Both the owner audit's `source_artifacts` and the certificate's fresh invoice (`input_sha256.fresh_invoice` → `PREFIX-AND-INVOICE.json` → `providers.gauge44.artifacts`) pin it.
- **The chain's own audit checks structure only.** `gauge44-slots-t16/code/audit_allcold.py` (lines 56–61) checks only each wrapper's `frame`, `terminal`, `vectors`, `terminal_vectors` and ranks.
- **Exactness is established only by the producer's own code.** `hpe_adapters_local.py` generates the wrappers and decodes them again with the same `clifford_exact` primitives, then writes `WRAPPER-CHECKS.json` (`PASS_RESIDUAL_WRAPPERS_ARE_EXACT`, sha256 `0ae753d6`).
- The fresh invoice pins that receipt's bytes, but no chain code reads its status or content.
- The cold replay re-runs the same producer, for example `gauge44-slots/construction/slots/verify_slots.py:103`.
- **#1781 (ours) left these records out** of its independent check. It covered the 2,896 child records and the 224 endpoint adapters.

## What this package checks

### Binding
- It reuses the merged #1781 binding, which our package imports with its digests pinned. The chain runs: selected record → frontier1094 certificate → capsule → owner audit → gauge44 word and programs.
- `RESIDUAL-ADAPTERS.json` must match the owner audit's own digest.
- The `split-centres` decoy, which has the same basename, differs by digest.
- The certificate's fresh invoice pins both `RESIDUAL-ADAPTERS.json` and `WRAPPER-CHECKS.json`.
- The global8 twin is byte-identical. This is informational only.

### Inventory
- **The keys.** The 36 keys are exactly the word's entering helpers ending at FULL (key = start) together with its hyperplane-ending helpers (key = start:final). That gives:
- 22 FULL-ended records;
- 12 cold hyperplane-end records (`0:U`);
- 2 entering hyperplane-end records (`2:999` and `3:998`).
- **The records.** Every record's frames, vectors and ranks match the word, and all 72 monomials are well formed.
- **The program events.** In both programs, every residual event is a `pre` or a `post`. Each of the 240 wrapped helpers has exactly one of each, with the right key. Its first event is the `pre` and its last is the `post`. Our open #1964, from another session, checks this placement as part of Lemma G's H1. Each orientation has 480 events, with widths {4: 112, 5: 64, 7: 60, 8: 244}.
- **Record fields.** Every record has exactly its expected field set.

### Two kernels
- **Kernel B** ([closed_form.py](closed_form.py), new, standard library only) computes amplitudes directly from closed forms:
- the frame operator from its closed formula;
- the monomial records from their closed forms;
- F_w from the √X product.

It works on sparse exact columns and evaluates no gate list.
- **Kernel A** is #1781's packed Gaussian-integer evaluator (`clifford.py`).

Both must find every identity exact on all 512 columns, with the global phase included.

### Calibration of kernel B
Before any record is judged, kernel B must pass three checks:
- **Frame operator.** Kernel B must reproduce #702's 14 frame-kernel digests and #1415's three operator digests of T_U ∘ T_E⁻¹. These cover frames of rank 1, 4 and 8 and the conjugate transpose.
- **Monomials and F_w.** On every column, kernel B's monomial and F_w must equal kernel A's: for 10 record monomials, and for F_w with w = 1, 4, 8 and 9. Without this check, those two conventions would be anchored only by the records under test.
- **Lean pairs.** Records `3:998` and `2:999` must be field-for-field equal to Exact.lean's pair 2 and pair 3. For `2:999`, #1611 had already shown this for this word.

## Results ([RESULT.json](RESULT.json))

| | Records | Exact (kernel B) | Exact (kernel A) | K (C gates per identity) |
|---|---|---|---|---|
| FULL-ended | 22 | 22 | 22 | 9 |
| Cold hyperplane-end `0:U` | 12 | 12 | 12 | 4 (8 records), 8 (4 records) |
| Entering hyperplane-end | 2 | 2 | 2 | 8 |

All 36 wrappers are exact.

### Controls
Each mutation below is rejected on every record it applies to:
- global phase +1;
- input character (`linear_Z4 += 2`);
- a CZ term in Q;
- an offset translation;
- swapped wrappers;
- width −1 and width +1;
- a transposed matrix (36/36 each);
- hyperplane-end records evaluated against FULL (14/14);
- a FULL-ended wrapper reused for a hyperplane end (2/2);
- another basis of the same span (2/2, while the same rows reordered pass);
- kernel B's own phase and width mutants (8/8).

The tests also reject the decoy, a flipped archive byte, a dropped `pre` event, an event read with the wrong key, and a wrong record vector. [test_package.py](test_package.py) has 17 tests.

## Overlap

Prior coverage, checked against the repository:
- **The Lean pairs.** `2:999` and `3:998` are the Exact.lean pairs (#593, with #1415's semantic binding). #1611 tied `2:999` to pair 3 for this word.
- **The cold rank-4 records.** The eight records `0:124`…`0:327` are byte-identical to records in earlier provider files. One example is `research/complex-sw4-frame8-validation`; the `community-k492-audit` copies match up to the terminal frame id. #702 checked that family's cold wrappers by amplitudes, for PR626.
- **The four cold hyperplane-end records `0:1000`…`0:1003`.** They are hyperplane cases of #1697's audit of all 511 cold hyperplane terminals. That audit used wrappers its own run generated, not this word's stored bytes.
- **What was producer-only.** The 22 FULL-ended entrance wrappers had only producer evidence.

What this package adds: all 36 are now checked from the **selected** word's bound bytes, by two kernels, and tied to its programs.

## What this does not show

- **The executor bridge.** It does not show that the emitted program applies these records with exactly these semantics (Lemma G's H1 and the label bridge; see #1964).
- **Other recipes.** The child and endpoint records are #1781's, and the identity adapters are not checked.
- **Lean.** There is no Lean proof for the 34 records outside the Exact.lean pairs.
- **κ.** There is no new κ.

## Reproduce

```sh
python3 -I -B research/frontier1094-residual-wrappers-seed62/verify.py # about 40 s; PASS_FRONTIER1094_RESIDUAL_WRAPPERS_EXACT
python3 -I -B -m unittest discover -s research/frontier1094-residual-wrappers-seed62 -p test_package.py
python3 scripts/research_package.py --package research/frontier1094-residual-wrappers-seed62 --check-only
python3.12 scripts/research_package.py --package research/frontier1094-residual-wrappers-seed62
```

Requires standard-library Python 3.10 or later. It was run on Python 3.12.15 and 3.14.8 (macOS 26.5, arm64).

## Open next questions

1. **Does the emitted executor apply the child, endpoint and residual records with exactly these semantics?**
- Why it matters: record-level exactness, here and in #1781, still needs the bridge rec(A, B) = record circuit.
- First step: replay one helper's full event sequence (`pre` wrapper, climbs, `post` wrapper) through an independent statevector executor and compare the result with C_[w] ⊗ I.
- Blocked by: an independent executor for the emitted program format.
2. **Should the chain check the exactness receipt's content, not just its bytes?**
- Why it matters: the fresh invoice pins `WRAPPER-CHECKS.json`, but nothing in the chain reads its status. This package's RESULT is an independent receipt that could be checked instead.
- First step: propose a status check, or this package's RESULT digest, at the next frontier1094 integration.
- Blocked by: a maintainer decision.
3. **Can Lean's `Clifford.check` kernel-check the 36 packed identities for this instance?**
- Why it matters: two of them are already Lean-proved, so the rest would turn this replay into a formal result.
- First step: emit the 22 FULL-ended identities as Lean `check` terms and time `lake build`.
- Blocked by: a Lean toolchain, which is not installed here.

## Credit

- **Lean packages:** Lemma A (#516) and Lemma G (#509), both by DreamingOfClouds and formalizing #484.
- **The producer of the residual wrappers and `WRAPPER-CHECKS.json`:** `hpe_adapters_local.py` (DreamingOfClouds, after PR #458 by chafreaky). It uses `clifford_exact.py` (icekylinx), whose exact phase lowering adapts PR425 (jacklightChen).
- **`HyperplaneEnd/Exact.lean`:** jeremykothe2 (#593).
- **Calibration sources:** #1415 (commdata2338) and #702 (BigCactusLabs).
- **The packed kernel and binding:** our #1781.
- **Related checks:**
- #1697 (navinor): all 511 cold hyperplane terminals;
- #1611 (bbMisk): binding helper 450's wrapper pair;
- #1964 (ours, open): Lemma G H1 extended labels.

Prepared by Sagar Thakrar (thakrarsagar) with substantial AI assistance from Claude (Anthropic) via Claude Code.
- Parallel read-only scout and judge subagents selected the obligation and traced the producer-only evidence.
- The closed-form kernel, the verifier and the controls were then written fresh.
- A post-PR adversarial review corrected the supplier trace (the fresh invoice pins the receipt's bytes) and the overlap, and added the kernel cross-check and further controls (second commit).

No claim was posted to Discussions. New material is Apache-2.0.
99 changes: 99 additions & 0 deletions research/frontier1094-residual-wrappers-seed62/RESULT.json
Original file line number Diff line number Diff line change
@@ -0,0 +1,99 @@
{
"binding": {
"archive": "research/community-frontier1094-audit/selected-replay.tar.gz",
"archive_sha256": "f1da80610bf33c58527f18ad63b4350d0b864e1d4486141ee17ac9bed4d529d6",
"canonical_word_gz_sha256": "16bc0ebb04b4be6c0ade74badc6a0c0cdba4215eda86e4ff9f13abf8ecebd219",
"canonical_word_payload_sha256": "4ddb29fb9bc6f27e0c0fbfcf19a2a93a3496ad900d0f6d89059d0305c9576c0c",
"certificate": "research/community-frontier1094-audit/ASSEMBLY.json",
"certificate_sha256": "1122cb3a575f8c788345fc1bf4c108ac7c3250fb81a5988671f5eb04b3ab3045",
"global8_twin_byte_identical": true,
"local_programs_gz_sha256": "241fde27d4e43f5a9cb6b128f9f76c5131a53d09f0371676163f914d0b0fef3a",
"owner_audit_sha256": "30bafff060afcfb3644e951214d66cfaffe58242881c652e28003958b42127cc",
"recipes_sha256": "d0768f2164d5f64a2fdbec422f1cfde4e400818792f1904ca541958377243731",
"record_or_checkpoint": "certificates/selected-result.json or certificates/checkpoints/frontier1094.json",
"residual_adapters_sha256": "383f40aadd00a4ac700dfd0d4c22d6c6c0be9a9863116e4d69c7acd23bafae0c",
"review_sha256": "cb0e10a8b284f176526c17177e91186ce09aa3e88b4f6acc561051daa072b0d0",
"slot_pins_sha256": "55dd229684febecb80638eff9c6b3986623e8539c6c956753b350209c3e0efe2",
"template_sha256": "9c4c51425251389ee9a747c465814cd07c64812c15b7aeef704dad7d65395a74",
"wrapper_checks_in_owner_audit": false,
"wrapper_checks_pinned_by": "certificate input_sha256.fresh_invoice -> PREFIX-AND-INVOICE providers.gauge44.artifacts (bytes only; no chain code checks its status or content)",
"wrapper_checks_sha256": "0ae753d6fa6e922d8b8180da602828f6cb3f444fcd124d019acee74e145c0836"
},
"calibration": {
"kernel_B_vs_kernel_A": "kernel B monomials (10, all columns) and F_w (w = 1, 4, 8, 9) equal kernel A",
"lean_pairs": "records 3:998 and 2:999 equal Exact.lean pair 2 and pair 3 field by field",
"pr1415": "kernel B reproduces the three operator digests of T_U o T_E^-1 (frame operator at ranks 1 and 8, and its conjugate transpose)",
"pr702": "kernel B reproduces all 14 frame-kernel digests"
},
"controls": {
"cz_term": "rejected 36/36",
"frame_basis": "other basis rejected 2/2; reordered rows pass 2/2",
"full_wrapper_for_hyperplane_end": "rejected 2/2",
"global_phase": "rejected 36/36",
"hpe_record_against_full_end": "rejected 14/14",
"input_character": "rejected 36/36",
"kernel_B_mutants": "rejected 8/8",
"producer_wrapper_checks": {
"frames": 22,
"hyperplane_end_pairs": 14,
"status": "PASS_RESIDUAL_WRAPPERS_ARE_EXACT"
},
"swapped_wrappers": "rejected 36/36",
"translation": "rejected 36/36",
"transposed_matrix": "rejected 36/36",
"width_minus_one": "rejected 36/36",
"width_plus_one": "rejected 36/36"
},
"evaluation": {
"K_histogram": {
"4": 8,
"8": 6,
"9": 22
},
"columns": 512,
"global_phase_included": true,
"kernel_A_exact": 36,
"kernel_B_exact": 36,
"records": 36
},
"inventory": {
"events": {
"backward": {
"post": 240,
"pre": 240,
"widths": {
"4": 112,
"5": 64,
"7": 60,
"8": 244
},
"wrapped_helpers": 240
},
"forward": {
"post": 240,
"pre": 240,
"widths": {
"4": 112,
"5": 64,
"7": 60,
"8": 244
},
"wrapped_helpers": 240
}
},
"helper_classes": {
"cold_full": 240,
"cold_hpe": 48,
"enter_full": 162,
"enter_hpe": 30
},
"kinds": {
"cold_hpe": 12,
"enter_hpe": 2,
"full_end": 22
},
"records": 36
},
"scope": "Record level: all 36 residual wrappers of the selected gauge44 word satisfy L F_w R T_E = T_end on all 512 columns with global phase (Lemma A exact part); not the executor bridge, a Lean proof or a new kappa",
"status": "PASS_FRONTIER1094_RESIDUAL_WRAPPERS_EXACT"
}
102 changes: 102 additions & 0 deletions research/frontier1094-residual-wrappers-seed62/SOURCES.json
Original file line number Diff line number Diff line change
@@ -0,0 +1,102 @@
{
"schema_version": 1,
"base_commit": "179089e4111a56a3bbf1f512410e7ef789380a71",
"files": [
{
"path": "certificates/selected-result.json",
"usage": "provenance",
"git_blob": "ae6a99f28aa468649eb66a361e764d769363bf11",
"sha256": "10b9e340fd8e354a37a0c4813c90e5de1dd42044b5f6aca4a3fc5a42b685fa32"
},
{
"path": "docs/research/agent-task-catalog.json",
"usage": "provenance",
"git_blob": "06c96f5579bb67ed1b86d334e78fbed7da6945f1",
"sha256": "6686ad58f1fbdbc87134a78901b60fd53a20d3984355e91defbbebc9e9fa8075"
},
{
"path": "docs/research/community-batch1099-integration.json",
"usage": "input",
"git_blob": "9d60e86c0e494d17bec5e28eda41b275c9ede64e",
"sha256": "684c50d1ac69863e1fcc0a34bf76c54bac576f6bbde79b3c67d8432a495c8c1c"
},
{
"path": "research/cold-wrapper-amplitude-audit/EXPECTED.json",
"usage": "input",
"git_blob": "64bb189e2e5c86da68f1b53bec8296a4220f583b",
"sha256": "64d9a891148c84bb6e5454ac6a02399a2abb3e553ee62f2719f8ed061eb8a501"
},
{
"path": "research/community-frontier1094-audit/ASSEMBLY.json",
"usage": "input",
"git_blob": "5fd676df060f3b450782fdf5ff5f6beb63a5eeb0",
"sha256": "1122cb3a575f8c788345fc1bf4c108ac7c3250fb81a5988671f5eb04b3ab3045"
},
{
"path": "research/community-frontier1094-audit/SELECTED-SOURCE.json",
"usage": "input",
"git_blob": "6a75493790b4eff9c4a98051ef8b01d6aad0d5a8",
"sha256": "79b3de5db3dbcb9829789ba58e9e45758926031e2a7d80667a08429d369be53b"
},
{
"path": "research/community-frontier1094-audit/selected-replay.tar.gz",
"usage": "input",
"git_blob": "c24daf067f9c543421d89ca2eeaa4eb78a57fb6f",
"sha256": "f1da80610bf33c58527f18ad63b4350d0b864e1d4486141ee17ac9bed4d529d6"
},
{
"path": "research/e8-lemma-a-lean/README.md",
"usage": "provenance",
"git_blob": "42355bfe904684da18f4ed1763dfcefb36eb36c8",
"sha256": "87a242f674783d6c553c450b7c5951f667f0f480316a557146449c5003c1d298"
},
{
"path": "research/e8-lemma-g-lean/PROOF.md",
"usage": "provenance",
"git_blob": "6b24da0ab861902a6a3f65d9599bb9c0f964aa1c",
"sha256": "2a7d84a83d34495dcf46ff9241ee3d7ec18ffc7d49c5476ab15fe0834046f2a3"
},
{
"path": "research/frontier1094-recipes-exact-seed76/CASES.json",
"usage": "input",
"git_blob": "e508b1224dedb4be9424cac58c361825c5a79810",
"sha256": "012493aa8d71a5f4e91fd82dd828daaab993efa36189a13b71b5c88eac536863"
},
{
"path": "research/frontier1094-recipes-exact-seed76/clifford.py",
"usage": "input",
"git_blob": "6f1e6a38febde7eb7a7341d317522e6650d84408",
"sha256": "d5e42ddabd141b7bfbd70fe13908f3c3a06aeccebeab90623596bfb1988b97fb"
},
{
"path": "research/frontier1094-recipes-exact-seed76/verify.py",
"usage": "input",
"git_blob": "8775d15e7473ee0b11ec47a07bc0d9cf0834bdd4",
"sha256": "7f66d587b6f73451b2b35941efe3bdd9271c83a88d64982a426f51b8c19336ef"
},
{
"path": "research/gauge44-slots-t16/code/audit_allcold.py",
"usage": "provenance",
"git_blob": "923bf7b675ca09f4d2633e7e53f16a1b23b6d5bc",
"sha256": "1467a2acb654e62552c69fb9a7c4c85f70a02120881d222313865058ab71a6cc"
},
{
"path": "research/gauge44-slots-t16/construction/slots/SLOT-PINS-k8.json",
"usage": "input",
"git_blob": "2702c368f4de0b671577aee9a0e2df41c1b27658",
"sha256": "55dd229684febecb80638eff9c6b3986623e8539c6c956753b350209c3e0efe2"
},
{
"path": "research/hpe-lean-semantic-binding/RESULT.json",
"usage": "input",
"git_blob": "c9169b5d5d483fc66dcb1ca5c27f2c2fa7124288",
"sha256": "9c0b3f39d9ef68ed8484f29fd9cb5d57e931580a882157afb593fb28839cbfdf"
},
{
"path": "research/hyperplane-end-lean/HyperplaneEnd/Exact.lean",
"usage": "input",
"git_blob": "1af33a206a9ba6c4826b3a1516d104e770380869",
"sha256": "eebf3a3982af7ab9be45777b272a6aaa523206fa7667ceb6df63ad6589fd8183"
}
]
}
Loading
Loading