Independent verification (no κ): #2435's words, 289,044-role owner map and endpoint reproduce; its change is frame-only (24 raised entrances), entrance delivery unchecked

[thakrarsagar]

Independent verification; no κ claim. This PR re-checks #2435 (Jeremy Kothe, head 324d7d35, conditional κ = 0.0016985113437, the newest top T74 claim) with code written independently of #2435. The checkers are this author's earlier #2387 fill_check80.py and #2315 words_check.py, used unchanged as libraries.

Everything checked reproduces:

• Words. Both words, recounted by two implementations, match the reported profile {1:2359, 2:632, 3:330, 4:120, 5:38, 8:8} and inventory {4:33, 5:9, 7:62, 8:90, 9:240}, with 434 helpers and 78 zero buffers.
  • All 46 warps pass: G invertible, K_G symplectic, Fourier rank 0.
  • The exact replay over ℚ⟨F⟩ passes forward and dual with all 658 inputs dirty: Y += X, with everything else and all 78 buffers restored.
• Owner map. OWNERS.json.gz (289,044 roles) passes rules R-a to R-g.
  • The 1,332 windows were rebuilt from 2,220 GLOBAL-TEMPLATE events and match Conditional κ = 0.001698511343697629: entrance raise of 24 copy buffers on the handoff T74 words #2435's PLAN.
  • X2 pending axes are disjoint, and the 9,129 distinct 73-column charts match SPEC actual_73column_sigmas.
• Recount equals Conditional κ = 0.001698511343697629: entrance raise of 24 copy buffers on the handoff T74 words #2435's SPEC, PROFILE and body.
  • Idle multiset {1:20824, 3:6695, 5:1, 48:25302}, global profile, 276,407 cross-replica roles, and rank mass 4,816,512 with deficit 23,680.
  • Physical stock: 66,748 under Improve conditional κ to 0.001697060296258836 with coherent handoffs and retimed hosting #2322's in-tableau rule, plus 78 buffers charged separately, gives 66,826.
• Endpoint. The root of the recounted profile is within ±1e-15 of the claimed κ (800-bit enclosures). The claimed κ also matches Conditional κ = 0.001698511343697629: entrance raise of 24 copy buffers on the handoff T74 words #2435's pinned outer/OUTER-CLAIM.json.
• Controls. 16 owner-map and word negative controls plus a flipped-coefficient replay control are all rejected.

Structural finding: #2435's change is frame-only. Its words differ from #2322's only in frames: 24 start_extra entrances, 24 A-gate frames and 48 B-gate frames per word. Registers, coefficients, warps, centres, ends and cleanup are byte-identical, and the script checks this.

• The scalar identity is therefore Improve conditional κ to 0.001697060296258836 with coherent handoffs and retimed hosting #2322's, unchanged.
• The one obligation specific to Conditional κ = 0.001698511343697629: entrance raise of 24 copy buffers on the handoff T74 words #2435 is that the parent hosting delivers those 24 copy buffers at a two-dimensional entrance frame. That is not checked here; Conditional κ = 0.001698511343697629: entrance raise of 24 copy buffers on the handoff T74 words #2435 covers it with its native tooling.

Related. #2442 shows this fill is not the pattern-integer optimum for these words. The optimum would add up to +5.8e-10 if realizable. That does not affect the correctness verified here.

Review. An adversarial AI reviewer ran about 20 additional mutations; every illegal one was rejected. It found the frame-only diff and three wording and hygiene issues: scope wording, a missing FAIL line on replay rejects, and an unpinned outer κ. All are fixed.

Checks. --check-only, the 3.10 grammar check and the research tests pass. Run on CPython 3.14.4; the registered 3.12 replay was not run.

python3 -I -B research/t74-2435-fill-independent-check/check_2435.py --fetch   # ~20 s
# PASS_T74_2435_FILL_INDEPENDENT roles=289044 idle={"1":20824,"3":6695,"48":25302,"5":1} physical=66826 charts=9129 controls=16 root=kappa_2435+-1e-15

Scope.

• Not checked: native lowering, invoice, precision tapes, outer ledger, bit supplier and admission. κ = C − 1/F − B is not re-derived.
• Base: main 0fdd7d82. Card: independent-moment, scope refined; the assignment is the independent-verification track, seed 83.

Credits.

• Conditional κ = 0.001698511343697629: entrance raise of 24 copy buffers on the handoff T74 words #2435 / Conditional κ = 0.001697194950901770: exact owner fill on #2322's handoff construction #2374 / Conditional κ = 0.001697065994108880: owner relocations on #2322's handoff construction #2349 (Jeremy Kothe) and Improve conditional κ to 0.001697147064845350 with whole-role augmenting paths #2363 (mason-hale).
• Improve conditional κ to 0.001697060296258836 with coherent handoffs and retimed hosting #2322 (Vlado Vojdanovski).
• Conditional κ = 0.001696247139058474: capped phase-cell fusion outer (PR1673) on the T74 supplier with anticipatory dirty borrowing (PR2226, 43 high-helper aliases) on the T74 pivot-566+554 supplier: conditional kappa 0.0016962471390584742… #2269 (Jeremy Kothe).
• Research: 80-helper anticipatory borrowing component and exact price rejection (no new kappa) #2226 (Cinna Davis).
• Improve conditional κ to 0.0016965151242727336 with a pure 44th dirty alias #2309 (Dugongue).
• Negative result (no κ): legal 400-helper owner fill of #2339's 80-alias T74 words reaches b = 0.00168654, 99.994% of a certified width-aware cap b < 0.0016867 — below #2269 #2387 / Independent verification (no new κ): record candidate #2269 T74 borrow43 — words, 437-helper fill and moment root reproduce; stale evidence, unpriced buffers, X2 open #2315 (this author): the vendored checkers, pinned.

Prepared with substantial AI assistance (Claude Code, Anthropic, Claude Opus 5.5). No independent human review is claimed. Apache-2.0.

🤖 Generated with Claude Code