Repository navigation
Conditional κ = 0.00167166528321877: seven triangle-hybrid slots (PR987 search) on the k8 leaver word - #1018
Merged
Merged
Conversation
…8 hybrids and frame re-optimisation on PR799's word
…K cells, U8 hybrids and frame re-optimisation on PR799's word
…R987 search) on the k8 leaver word, verified on a checked CrocSwap#1005 baseline
This was referenced Oct 13, 2026
Mister-Meeseeks
pushed a commit
that referenced
this pull request
Oct 13, 2026
… new kappa) Standard-library re-implementation of #858's whole-helper hosting rules, reading the frozen OWNERS/PLAN/SPEC of every open seven-slot k8 T16 claim (#1022-#1087) as pinned git blobs. Windows are rebuilt from main's vendored source-event stream; widths from the #1049-pinned word. All 15 maps and the shared #1018 T8 map pass; 11 targeted negative controls. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
This was referenced Oct 13, 2026
gvconly
added a commit
to gvconly/integer-mult-bounds
that referenced
this pull request
Oct 13, 2026
…omoted word, perfectly packed (κ 0.001675338871723469) Fourth layer on the CrocSwap#1005/CrocSwap#1018 chain: the ten low-ended helpers 500, 501, 572-575, 580-583 of research/gauge44-slots-unpromo-arcflow's word get the seven-dimensional end frame their cones force (the join of the minimal end-of-B frames of the helper and its cone registers), the free gate frames are re-optimised exactly (PR864 model, HiGHS MILP), and the whole-owner map is rebuilt for the new inventory {4:40, 5:48, 7:40, 8:112, 9:240}: the windowed LP bound is the perfect packing and the liftable MILP attains it, idle remainder {48:2460} with no small remainder. Strict replay, independent price, recount and the layered verifier (three rejected mutants) all pass. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01KBSZTxNtzYkGa6MNbKajRo
gvconly
added a commit
to gvconly/integer-mult-bounds
that referenced
this pull request
Oct 14, 2026
…g optimum of the seven-slot k8 word New package research/gauge44-slots-arcflow on research/gauge44-slots (CrocSwap#1018 chain). Only the owner map changes: idle remainder {1:1214, 2:1, 32:1, 48:2443} over 34,560 whole-helper roles, found by an exactly liftable pattern/time MILP at the time-free LP bound and a Konig edge-colouring lift. Strict replay, admission and bit-supplier controls, independent price and a standard-library recount pass. Prepared by G. V. Chudnovsky with Anthropic Claude assistance. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01KBSZTxNtzYkGa6MNbKajRo
gvconly
added a commit
to gvconly/integer-mult-bounds
that referenced
this pull request
Oct 14, 2026
…omoted word, perfectly packed (κ 0.001675338871723469) Fourth layer on the CrocSwap#1005/CrocSwap#1018 chain: the ten low-ended helpers 500, 501, 572-575, 580-583 of research/gauge44-slots-unpromo-arcflow's word get the seven-dimensional end frame their cones force (the join of the minimal end-of-B frames of the helper and its cone registers), the free gate frames are re-optimised exactly (PR864 model, HiGHS MILP), and the whole-owner map is rebuilt for the new inventory {4:40, 5:48, 7:40, 8:112, 9:240}: the windowed LP bound is the perfect packing and the liftable MILP attains it, idle remainder {48:2460} with no small remainder. Strict replay, independent price, recount and the layered verifier (three rejected mutants) all pass. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01KBSZTxNtzYkGa6MNbKajRo
gvconly
added a commit
to gvconly/integer-mult-bounds
that referenced
this pull request
Oct 15, 2026
…nd ten deepened ends, perfectly packed (κ 0.001675784247432857) Fifth package on the CrocSwap#1005/CrocSwap#1018 chain, a sibling of research/gauge44-slots-deepen-arcflow on the slot word. Pricing the frame moves against perfect packing shows the deepened word over-converted: a restored entrance costs 3.7e-9 in kappa and removes a rank-1 remainder worth 4.3e-9, so once packing is perfect every further restoration is a loss. A joint integer programme over conversion counts and time-free bin patterns gives fourteen restorations and ten deepenings; the exact frame re-optimisation prices the two entrances left promoted cheapest when they are two of {603, 605, 606, 607}. Recipe, standard-library applier, four rejected mutants and verifier in construction/balance; owner map idle {48:2463}; strict replay, independent price, recount, CLAIM.json (five packages, caps and realised fractions) and the whole-head driver extended to five packages with a fresh cold receipt (22 stages) in wholehead/. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01KBSZTxNtzYkGa6MNbKajRo
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Conditional κ = 0.00167166528321877. That is +2.78×10⁻⁷ over #1005 (0.001671387264390146) and +9.70×10⁻⁶ over the merged #942. The complex side binds; PR589's bit supplier is unchanged.
Change: seven more triangle-hybrid slots on #1005's k8 word. The slot search and the first two slots (X = 444, 446) are rohanarun's #987. Running #987's search over k8's copy gates found five more legal slots (X = 236, 238, 434, 436, 439); a re-search on the seven-slot word finds none further. The mechanism is #869's triangle-hybrid B delivery.
construction/barch/build_tri.pyandrun_controls.py, unchanged (byte-identical), with the new recipeconstruction/slots/RECIPE-SLOTS-k8.json. Inputdb2c7cfb…, outputc2b29430…, widths unchanged.drop_restore,restore_full,drop_prereadanddonation_after, applied globally and to each of the seven slots. All are rejected as pinned inSLOT-PINS-k8.json.Verification
This follows the repository's explicit baseline-reuse pattern (as
research/complex-t37-owners/verify.py --baseline799).construction/slots/verify_slots.pytakes a completed #1005 chain replay and rechecks it: the complete step receipt (21 steps, all exit 0), the final word against #1005's frame-recipe pin, the canonical, program and Clifford locals againstCHAIN-PINS-k8, and all five layer-end control counts. It then rebuilds only the slot layer with its controls, the native locals of the new word (native, ledgers, local-replay, wrappers, phases),audit_local, and the strict global replay.From a fresh clone of commit eac7547 (Windows, CPython 3.11.11):
verify_manifest.py: PASS, 172 filesverify_slots.py --baseline <completed #1005 chain replay>: steps parent-manifest (1 s), slots (0 s), slot-controls (16 s), native (1 s), ledgers (1 s), local-replay (8 s), wrappers (2 s), phases (118 s), audit-local (1 s), strict (20 s); PASS_SLOTS_ON_CHECKED_K8_BASELINE 0.00167166528321877 (2 m 48 s total)The inherited
evidence/, proof notes and README body below the new section are #1005's files, retained unchanged as parent evidence. The current receipts for this package areRESULT.json(strict replay),construction/slots/SLOT-PINS-k8.jsonandverify_slots.py's output.Scope
Conditional on PR645's premises and PR589's bit-supplier assumptions; a finite exact construction. No unconditional theorem or Lean certification is claimed.
Credits
Prepared with Claude Code (Anthropic).