Skip to content
Open
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
1 change: 1 addition & 0 deletions research/p104-reroutes-on7623/EXPECTED.json
Original file line number Diff line number Diff line change
@@ -0,0 +1 @@
{"kappa_prefix": "0.0036667343458950"}
15 changes: 15 additions & 0 deletions research/p104-reroutes-on7623/NOTICE
Original file line number Diff line number Diff line change
@@ -0,0 +1,15 @@
Apache-2.0. DreamingOfClouds with Claude Opus 5.5 (Anthropic), via Claude Code.

The base word is our #7623 (14 loan + stationary-snapshot instances on our #7621), which uses SYLforge's device of
#7616 and #7620 and inherits #7620's composition: Ali Romman's #7605, our #7611 stay drops, Clear1nsight's #7612
contact rechart, FanPUYen's #7604, SYLforge's #7603, ToikY's #7546/#7544 and rohanarun's signed #7113 caller,
and the courier lineage of our #7615/#7617 and the crowd's courier pattern (Clear1nsight's #7244/#7274/#7314/#7334/#7481,
rohanarun's #7480 and #7277, CubeJerry's #7577, petalas's #7326). #7627 (Clear1nsight) is cited only as a price point.

Vendored data (unmodified, data only): data/base-7623-word.json.gz, data/invoice-7623.json and
data/affected-pairs-7623.json are our #7623's data/word.json.gz, data/invoice-new.json and data/affected-pairs-loan.json.

check.py is our own code. Its replay, crossing and column-test sections are taken from our #7623 package; its column
tests follow the description of the affected-pair tests in #7620's README and native runner. The CLEAR certificate
check implements the acceptance rule of #7620's verify_certificates (read, not run). No other author's code was run;
their words are read as data.
108 changes: 108 additions & 0 deletions research/p104-reroutes-on7623/README.md
Original file line number Diff line number Diff line change
@@ -0,0 +1,108 @@
# CONDITIONAL κ = 0.0036667343458: seven stop edits and a joint hub re-settle on #7623 (premises: the same as #7623's)

**κ = 0.00366673434589504230429298346749029694192150…, ×1.00118271 (+0.118271%) over our
[#7623](https://github.com/CrocSwap/integer-mult-bounds/pull/7623)** (κ 0.0036624027915539817894…), and ×1.00027363 over
[#7627](https://github.com/CrocSwap/integer-mult-bounds/pull/7627) (κ ≥ 0.003665731276). The official `evaluate_recursive_power`
bracket (256 bits) is [0.0036667343458, 0.0036667343459]. The claim 0.0036667343458 is below the exact crossing 0.00366673434589504.

> **Premises: conditional on the same framework assumptions as #7623.** Those are exactly #7620's premises: its five framework
> assumptions (`native_all_columns`, `owned_erasure`, `compiler`, `precision_tape`, `analytic_recovery`) and its fixed allocation
> of 17·68 = 1,156 owned banks, with its CLEAR reading over #7623's **180 CLEARs**. This word adds no owned lifetime and no new kind
> of event. The 180 owned lifetimes (field, INIT chart, CLEAR chart), the 47 owned MOVEs, peak 50 and 68 labels are #7623's. The
> CLEAR *contents* do change (156 of 180), because the hub content they snapshot is re-derived. Each of the 180 CLEARs therefore carries its
> own exact certificate under #7620's `verify_certificates` rule (`data/clear-certificates.json`). If #7620's premises are refused,
> this word is void with #7620, #7621 and #7623.

## The edits

Base: #7623's word (`data/base-7623-word.json.gz`, 77,277 events). New word: `data/word.json.gz`, 208,738 events. Every edit is listed
in base-word coordinates in [data/edits.json](data/edits.json). That covers 23 deletions and 12 insertions on the interior events (every
MOVE/INIT/CLEAR, and every ADD/COPY/SCALE at a chart of dimension 1..8), plus `initial` for the two rebirths. Then come the re-derived hubs.

| id | register | kind | route #7623 → new | new ops |
|---|---|---|---|---|
| E1 | helper 457 | rebirth + swap canceller (TR1) | 3→281→966→218→127 ⇒ born at 281 | `242 += -1*457` @281 |
| E2 | helper 461 | rebirth + swap canceller (TR1) | 3→285→994→201→127 ⇒ born at 285 | `10 += -1*461` @285 |
| E3 | helper 382 | re-route at an existing co-presence | drops 24: 3→849→… | `553 += 1/2*383` @24 (replaces 4 ops) |
| E4 | helper 490 | re-route at an existing co-presence | drops 725: 594→883→… | `1427 += -1*296` @725 (replaces `490 += 296`) |
| E5 | X main 44 | re-route at an existing co-presence | drops 849: 495→127 | `382 += -1*816` @849 (replaces `382 += 44`) |
| E6 | Y main 176 | re-route (twin pickup) | drops 1065: 449→928 | `176 += -1/4*1000`, `176 += -1*236` @928 (replace `176 += -1/2*379` @1065) |
| E7 | Y2 main 902 | stop deletion, no new op | drops 595: 3→618 | none (4 donor deliveries deleted) |
| hub | {0} chart and V | joint exact re-settle | n/a | every ADD/SCALE at dim 0 and dim 9 re-derived |

**Why each kind is legal.**
- **TR1 swap canceller (E1, E2).** Helpers 457 and 461 were born at {0} and moved straight to their dim-3 first chart. Each is now
born at that chart; its first MOVE is deleted. While its row is still exactly e_h, one canceller op `j += -1*h` runs at that chart,
between registers that are both there. That is TR1's depth-1 swap: the token e_h now leaves on j. j = 242 is a {0}-born helper that
ends at V, and j = 10 is a kept X main. Its content is settled by the V mix, so no paid stay carries the token. Each helper's target
rank goes 9 → 6.
- **Re-route at an existing co-presence (E3–E6).** The register drops one interior stop with all its ops there, and its two MOVEs merge
into one strictly ascending MOVE. The content the stop delivered is re-made by one or two new ADDs between registers that already
share that chart at that moment in #7623's schedule. Each new op's legality is a chart check in the exact replay. For E5 the receiver
382 ends at V, so the extra X content stays inside the V span and is absorbed by the V mix, and the {0} part by the {0} prefix. For E6
the X content delivered to 176 is identical; the difference is {0} content plus two Y2 entries absorbed downstream.
- **Deletion with the hub re-settle (E7).** Main 902's stop at 595 is donor-only (it delivers to owned 1532 and to helpers 1479, 1255
and 1495, which end at V). The four deliveries are deleted with no replacement. A positive control shows that the bare deletion
on #7623 already leaves every final row unchanged, because the deliveries cancel downstream.
- **Joint hub re-settle.** Every ADD/SCALE at the {0} chart (the prefix among helpers born at {0}) and at the full chart V (the mix of
registers ending at V) is solved jointly and exactly. The solution makes every final row equal #7623's and is factored into
ADD/SCALE events with 5-smooth coefficients (61,251 → 192,723 such events).

Exactly the 7 edited registers change chains. Five new MOVE pairs appear, all strictly nested: (3,618), (3,849), (449,928), (495,127),
(594,883). Two leave: (24,849), (595,618).

## Price

Native paid delta per call {1:−2, 2:−3, 3:−5, 5:+1, 6:+2}; helper target delta per call {6:+2, 9:−2}. Over 17 calls: paid
{1:−34, 2:−51, 3:−85, 5:+17, 6:+34}, targets {6:+34, 9:−34}. Paid positions 110,145 → 110,026; rank surplus S = 1,424 unchanged.
Official tools (`price.py`):
- `compare_recursive_invoices` against #7623 gives `uniformly_better`, strictly negative for r^(1−C).
- `evaluate_recursive_power` (256 bits) gives `new_cost_lower` at C = 0.0036667343458, 0.0036624027915 (#7623) and
0.003665731276 (#7627), and `new_cost_higher` at C = 0.0036667343459.

Ordinary bill (17 calls, the repository `independent.py` formula with its call count set to 17): 786,433,632 for #7623 → 867,014,040
for this word, below 10^9. The re-derived hubs triple the word length (77,277 → 208,738 events), and that is where the increase comes from.
Coefficients are 5-smooth. The replayed grid moves from #7623's (2-adic exponent 25, odd lcm 1,125, 40 numerator bits) to (20, 675, 34),
and non-dyadic literal events go from 5,903 to 12,312.

## Checks

```sh
python3.12 -B scripts/research_package.py --package research/p104-reroutes-on7623
```
`VERIFY.json` registers six commands:
- `core`: words, interior edit list, exact replay and legality of both words, rows, end charts and owned data, routes, coefficients,
invoice and κ.
- `certificates`: all 180 CLEAR certificates, with 2 rejecting controls.
- `controls`: 14 rejecting controls and 3 accepting.
- `columns1` and `columns2`: the affected-pair column tests.
- `price.py`.

**Affected pairs.** #7623's census is 964 pairs. This word changes 7 routes, and the census becomes **981 pairs**: +17, the old and new
route pairs of the changed registers that were not already counted, with none dropped and every new MOVE pair inside. `check.py` runs 512
forward and 512 adjoint Gaussian columns, signed Pauli action, absolute phase, exact ascent rank and a wrong-phase rejection on all 981
pairs, in two halves.

Independently of the package, the word passed our exact pricer, an independent exact row replay with chart legality, a replay modulo 2^89−1 and 2^107−1,
the repository's `independent.py` native audit (bad_inputs 0) and caller identity (bad_outputs [], 17 native calls), and an interval
gate on the κ bracket, with 6 rejecting controls.

## Conditional premises and what is not rerun

The claim carries exactly #7623's premises (above), over the same 180 CLEARs with new, individually certified contents, and the signed
#7113 caller and BIT premise of the base. This package does not rerun #7620's full native/signed 17-stage application, catalogue, outer
predicates or precision/tape counts. The header, end rows, owned lifetimes and labels are unchanged, but the coefficient grid and event
count are not (see Price), so the precision/tape figures quoted for #7620 do not carry over verbatim. No upstream maintainer has reviewed
the package.

## Credit

The base is our [#7623](https://github.com/CrocSwap/integer-mult-bounds/pull/7623) on [#7621](https://github.com/CrocSwap/integer-mult-bounds/pull/7621),
which uses SYLforge's loan + stationary-snapshot device ([#7616](https://github.com/CrocSwap/integer-mult-bounds/pull/7616),
[#7620](https://github.com/CrocSwap/integer-mult-bounds/pull/7620)). Signed caller: rohanarun's #7113. See [NOTICE](NOTICE).
DreamingOfClouds with Claude Opus 5.5 (Anthropic), via Claude Code. Apache-2.0.

## Open next questions

**The premises.** A ruling on #7620's inherited assumptions and its CLEAR reading (`owned_erasure`, clear-form-gap Q1) decides #7620,
#7621, #7623 and this word together. See [TASK.json](TASK.json).
30 changes: 30 additions & 0 deletions research/p104-reroutes-on7623/SOURCES.json
Original file line number Diff line number Diff line change
@@ -0,0 +1,30 @@
{
"schema_version": 1,
"base_commit": "e54523517d0f03e58352e647c403a5af3e525a57",
"files": [
{
"path": "certificates/selected-result.json",
"usage": "provenance",
"git_blob": "75d77b328ea02ded535e7f23aee5b1e55d0bb1a8",
"sha256": "154c3d30455d778fd3343b6b4851f0fa4665f0f7625b0adb3675e2bafbbf6bf1"
},
{
"path": "docs/research/agent-task-catalog.json",
"usage": "provenance",
"git_blob": "3404fbdd5e7529a0674eef26114b5071e4586cec",
"sha256": "ead6cdc9d0af6d58e8a49744d8ec8c7c15e13d2ba93c276c434bcf9b51b93a81"
},
{
"path": "scripts/compare_recursive_invoices.py",
"usage": "input",
"git_blob": "fda2a852e9937e42fe40df0707694465ca7c1b48",
"sha256": "3930c22c11da7bc6042cfeb6e459ecb06e4478430037bee3469d034353855fc6"
},
{
"path": "scripts/evaluate_recursive_power.py",
"usage": "input",
"git_blob": "ca50f8021fc3238a58fd420383f4619a8d008ace",
"sha256": "52b20f41fd0bb55fa88dccd6015af2ce358917817b9506720f3e60ce681cc3ad"
}
]
}
117 changes: 117 additions & 0 deletions research/p104-reroutes-on7623/TASK.json
Original file line number Diff line number Diff line change
@@ -0,0 +1,117 @@
{
"schema_version": 1,
"task_id": "p104-side-correction",
"priority_revision": 8,
"base_commit": "e54523517d0f03e58352e647c403a5af3e525a57",
"original_card": {
"id": "p104-side-correction",
"question": "Can joint helper bases and a legal causal split improve the complete P104 productive/paid invoice?",
"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/p104-helper-subspace-splits/README.md",
"research/dyadic-helper-channel-capacity/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": "An exact arbitrary-input identity, source/target frame availability and restoration at the chosen cut, plus the complete channel+correction+idle/hosting bill or an explicit favorable bound. The native channel alone is not a supplier. 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": "Start with the selected #5599 emitted word and review pending #5718 distinct input/output bases plus #5717/#5719 coverage before claiming work. Pick one uncovered source word, helper group and cut region. Prove the arbitrary-input relation and legal entry/exit corrections, then compare full target and paid profiles. Reject if completion, owned lifetimes or consumer costs repay the gain. A static pairing rank alone is not a physical split.",
"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": [
"P104",
"helper basis",
"helper split",
"5718",
"dyadic capacity"
],
"existing_checks": [
{
"scope": "PR #4133: literal-slot defect",
"evidence": "research/p104-literal-slot-defect/README.md",
"remaining": "Independently validate the stated scope or change its producer/consumer premise and pay the complete boundary.",
"status": "author-checked",
"source_commit": "88666f1dee93fbd85b31fd5d6f91a219e111b5f4",
"selected_record_sha256": "bb2eaa58647b353f88d0aec2eafb9d30df1be53589231b23e12ceef3e4bad129"
},
{
"scope": "PR #4136: uniform commutation gap",
"evidence": "research/p104-uniform-commutation-gap/README.md",
"remaining": "Independently validate the stated scope or change its producer/consumer premise and pay the complete boundary.",
"status": "author-checked",
"source_commit": "281ce601a8be9db470de630759179c23e1dfaa88",
"selected_record_sha256": "bb2eaa58647b353f88d0aec2eafb9d30df1be53589231b23e12ceef3e4bad129"
},
{
"scope": "PR #4140: uniform encoding gap",
"evidence": "research/p104-uniform-encoding-gap/README.md",
"remaining": "Independently validate the stated scope or change its producer/consumer premise and pay the complete boundary.",
"status": "author-checked",
"source_commit": "de25c6525aa866962309df5d107b32d55e6e5628",
"selected_record_sha256": "bb2eaa58647b353f88d0aec2eafb9d30df1be53589231b23e12ceef3e4bad129"
},
{
"scope": "PR #4153: source-chart shortcut exhausts its optimistic allowance",
"evidence": "research/p104-source-chart-shortcut/README.md",
"remaining": "Independently validate the stated scope or change its producer/consumer premise and pay the complete boundary.",
"status": "author-checked",
"source_commit": "9123241fea53df77e51ffe30d9aca1625d4bf3a8",
"selected_record_sha256": "bb2eaa58647b353f88d0aec2eafb9d30df1be53589231b23e12ceef3e4bad129"
},
{
"scope": "PR #4154: fiber-tag seam mismatch",
"evidence": "research/p104-fiber-tag-seam/README.md",
"remaining": "Independently validate the stated scope or change its producer/consumer premise and pay the complete boundary.",
"status": "author-checked",
"source_commit": "4ffcdaee5d6255a9cea1fe53a5424435068089d0",
"selected_record_sha256": "bb2eaa58647b353f88d0aec2eafb9d30df1be53589231b23e12ceef3e4bad129"
},
{
"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": "Exact #7623 word (our loan + stationary-snapshot word on #7621). Seven stop edits plus one joint hub re-settle: two TR1 swap-canceller rebirths (helpers 457 and 461 are born at their dim-3 first chart; 242 += -1*457 and 10 += -1*461 while their rows are still e_h), two helper re-routes at existing co-presences (382 drops chart 24: 553 += 1/2*383; 490 drops chart 725: 490 += 296 becomes 1427 += -1*296), two main re-routes at existing co-presences (X main 44 drops chart 849: 382 += 44 becomes 382 += -1*816; Y main 176 drops chart 1065: its pickup from 379 becomes 176 += -1/4*1000 and 176 += -1*236 at 928), and one plain main stop deletion (Y2 main 902 drops donor-only chart 595). Every ADD/SCALE at the {0} chart and at V is then re-derived jointly and exactly so that all final rows equal #7623's. Rows, end charts, owned MOVEs (47), the 180 owned lifetimes, peak 50 and 68 labels are #7623's; the 180 CLEAR contents change with the hub content and each carries an exact certificate under #7620's rule. Conditional on the same framework assumptions as #7623.",
"scope_change_reason": "Removes seven paid stops from #7623's word by re-routing their content through registers that already meet there (or, for 902, dropping a donor-only stop) and re-settling both hubs exactly; each edit was found by an exact screen over all interior stops of #7623 and checked by exact replay before it was stacked.",
"bottleneck_link": "Reduces the complete signed paid invoice by {1:-34, 2:-51, 3:-85, 5:+17, 6:+34} and moves two helper targets from rank 9 to rank 6 ({6:+34, 9:-34}) over the 17 native calls; paid positions 110,145 -> 110,026, rank-weighted S 1424 unchanged.",
"budget_minutes": 120,
"stop_rule": "One literal word with exact arbitrary-input row equality, legality, the interior edit list, exact certificates for all 180 CLEARs, all affected-pair column tests (981 pairs), complete invoice and a certified crossing.",
"outcome": "completed",
"evidence": "README.md, data/edits.json, check.py, price.py, VERIFY.json.",
"author": "DreamingOfClouds",
"ai_assistance": "Claude Opus 5.5 (Anthropic), via Claude Code.",
"open_questions": [
{
"question": "Do #7623's premises (#7620's five inherited assumptions native_all_columns, owned_erasure, compiler, precision_tape, analytic_recovery, its 180-CLEAR reading and its fixed 17*68 bank allocation) hold?",
"why_it_matters": "The claimed kappa holds only if #7623's does. This word adds no owned lifetime and no new kind of event; its 180 CLEARs are the same lifetimes as #7623's with new contents, each certified exactly.",
"first_step": "Run check.py core and check.py certificates.",
"blocked_by": "A ruling on #7620's premises (owned_erasure, clear-form-gap Q1)."
}
]
}
Loading
Loading