Perception · W5 — Multi-Formation Count Bridge

W5: 2026-04-27 to 2026-05-03.
Canonical bump: v2.3 → v2.4 through CV-1.5, CV-1.5.1, and CV-1.5.2.
The week’s center of gravity: multi-formation count theory finally gets its first Cat A bridge.

Start here#

W4 resolved the old Critical-3. W5 used the newly cleared space to do harder accounting: multi-formation structure, count bridges, proof hygiene, demotions, and open-problem registration.

The headline is T-L1-F:

Under the L1-J regime, the hard-bar count of the aggregate field equals the active-slot count of the multi-formation state, with an explicit labeled bijection from active slots to dominant persistence bars.

That is the first multi-formation Cat A theorem in SCC.

What makes the week readable is not just that a theorem was added. It is that the week also kept track of what did not become theorem-grade: T-σ-Theorem-4 was demoted, V5b-T' was withdrawn, and two HIGH open problems were registered instead of being papered over.

This week in five lines#

• T-L1-F entered canonical as Cat A conditional. This closes the L1-A through L1-L working chain and gives SCC a labeled bridge between active slots and aggregate-field persistence.
• L1-M was drafted as the soft-count companion. It is Cat-B sketched, with a clean audit path toward CV-1.6.
• The σ-framework expanded, then corrected itself. CV-1.5 added supporting structures; CV-1.5.1 added multi-formation static σ entries and Commitment 16.
• T-σ-Theorem-4 was demoted from Cat A to Cat B. NQ-187 found a numerical contradiction; the demotion is part of the result, not an embarrassment hidden in the footer.
• Two HIGH open problems were registered. OP-0008 and OP-0009 make the next layer of uncertainty explicit.

Scoreboard#

Footnote on the open-problems row: the W4-close active set was 7 (1 High = OP-0005, 4 Medium = OP-0010..0013, 3 Low = OP-0020..0022); OP-0004 had been retracted and OP-0001/0002/0003 were resolved, clarified, or sidestepped. The W5-close active set is 11 (4 High = OP-0005, OP-0006, OP-0008, OP-0009; 4 Medium = OP-0010..0013; 3 Low = OP-0020..0022). The +2 HIGH delta reflects the CV-1.5.1 (W5 Day 4) registrations of OP-0008 (σ^A K-jump non-determinism) and OP-0009 (Multi-Formation Ontological Foundations, with 7 sub-items). OP-0006 (Boundary precision) was promoted from Medium to High during W5.

The raw increase matters less than the shape of the increase. W5 added theorem-grade multi-formation structure while also registering the places where the theory is still underdetermined.

The week’s arc#

Day 1: σ support enters canonical#

CV-1.5 merged five σ-framework support entries:

• T-σ-Lemma-1: irrep decomposition is well-defined.
• T-σ-Lemma-2: nodal-count properties.
• T-σ-Lemma-3: Goldstone-saturation identity.
• T-σ-Theorem-3: closed-form σ at the uniform state.
• T-σ-Theorem-4: first-pitchfork leading-order σ, later demoted.

The important process detail: the merge was immediately followed by two audit rounds. Round 1 caught three numerical errors. Round 2 caught eleven structural-completeness issues. This made W5’s default discipline clear: canonical merge is not the end of verification.

Day 2: multi-formation σ opens#

The moderate G3 plan expanded into a ten-cycle LSW analysis cascade. The useful outputs were not a single theorem but a map:

• STRICT per-formation pool: $\alpha = -0.069$, Cat A verified.
• Standardized α-window LSW plateau: $\alpha = 0.25$ to $0.30$, Cat A.
• Hessian-based $\sigma_{\mathrm{multi}}^A(t)$ implemented, Cat B.
• K-jump timing statistics: $\Delta t \propto t^{1.315}$, LSW-consistent.
• 3D extension structurally works at small scale, but statistics are not yet enough.

This was a working-theory expansion day, not a clean canonical-promotion day.

Day 3: dynamic σ exposes non-determinism#

The dynamic trajectory layer $\sigma_{\mathrm{multi}}^A(t)$ made a new problem visible.

At merger times, post-merger σ is not determined by pre-merger σ alone. One also needs merger geometry: which slots merge, their centroids, orientation alignment, and the relaxation path. That became OP-0008, registered HIGH.

The key move was honest self-downgrade: a claim first phrased as Cat C asserted was immediately revised to Cat C conjectured.

Day 4: CV-1.5.1 and the red finding#

CV-1.5.1 added the static multi-formation σ layer:

• T-Commitment-14-Multi-Static.
• T-σ-multi-A-Static.
• T-σ-multi-D-Static.
• T-σ-Multi-1 as a Cat B target.

It also added Commitment 16: K-status Two-Tier Decomposition:

$K = (K_{\mathrm{field}}, K_{\mathrm{act}}).$

$K_{\mathrm{field}}$ is the architectural cap. $K_{\mathrm{act}}$ is the dynamics-emergent active stratum index. This resolves OP-0009-K, the K-status sub-item that had been ambiguous across several incompatible usages.

Then NQ-187 produced the week’s red finding: numerical sweeps on $D_4$ free-BC contradicted T-σ-Theorem-4’s leading-order ratio. That forced the Cat A → Cat B demotion.

Day 5: reconciliation#

Day 5 was deliberately smaller. Its job was not to grow the theory; it was to stop the theory from accumulating stale working files.

What happened:

• 47 working files were checked.
• 15,805 lines were accounted for.
• 9 retractions were documented.
• The CV-1.7 parking lot was created for unaudited clusters.

This is the production lesson of W5: high-volume weeks need reconciliation days, or the working directory starts lying about what is real.

Day 6: T-L1-F promotion#

T-L1-F entered canonical at CV-1.5.2.

The theorem is conditional, but not vague. It depends on the L1-J regime package P0-P11: terminal-death convention, deterministic tie rule, active mass, disjoint active neighborhoods, boundary collar, background suppression, birth height, decay-to-cut, tightened H6, perturbation control, residual suppression, and a margin ledger.

The payoff is the labeled bijection:

$\mathcal A_{\mathrm{bar}}: A^\varepsilon \to \mathrm{Bars}_0^{\mathrm{term}}(U;G).$

Each active slot gets exactly one dominant terminal-death bar, born at its primary representative. This gives multi-formation persistence a label structure it did not previously have.

Day 7: L1-M soft-count companion#

L1-M was drafted as the controlled soft-count companion to T-L1-F. Under the same regime plus an envelope class $\Phi_{\mathrm{res}}$, it bounds the soft count against the active-slot count:

The useful surprise: the planned separate "edge-band control" hypothesis collapses into the existing L1-J constants when $\tau < \tau_*$. The hypothesis package is cleaner than expected, but the proof still needs L1-M-AUDIT before Cat A promotion.

Deep dive: what T-L1-F actually gives#

Take a multi-formation state

$\mathbf u = (u^{(1)}, \ldots, u^{(K_{\mathrm{field}})}).$

Some slots are active; some are negligible. Aggregate them into a single field:

$U(\mathbf u) = \sum_j u^{(j)}.$

Now count two things:

1. Active-slot count: $K_{\mathrm{act}}^\varepsilon(\mathbf u)$, the number of slots whose mass exceeds the threshold.
2. Hard-bar count: $K_{\mathrm{bar}}^{\ell_{\min}}(U;G)$, the number of long enough terminal-death $H_0$ superlevel persistence bars in the aggregate field.

T-L1-F says these counts agree under P0-P11, and more importantly, that the agreement is labeled.

That label is the theorem’s real content. Before T-L1-F, SCC could count active slots and it could count aggregate-field topological features. It could not canonically say which topological feature came from which slot. T-L1-F supplies that correspondence.

Non-claims are explicit:

• It does not say the equality holds globally.
• It does not solve K-selection.
• It does not solve OP-0008.
• It does not prove the soft-count version; L1-M is the companion.
• It does not promote reservoir theory or claim $\sigma_{\mathrm{rich}}$ sufficiency.

This is why "conditional Cat A" is the right status: theorem-grade within a named regime, no silent universalization.

Reclassifications#

T-σ-Theorem-4: Cat A to Cat B#

T-σ-Theorem-4 was merged on Day 1, then demoted on Day 4.

Why:

• NQ-187 measured the scaling exponent $p \approx 1.03$ at $L = 16$ on $D_4$ free-BC for $\Delta\mu = \mu_1 - \mu_0$ vs $\epsilon$ (canonical predicts $p = 2$), with $\mu_0/\epsilon \approx 1$ and $\mu_1/\epsilon \approx 2$ — i.e. the discrete-grid amplitude ratio reads $A_2/A_1 \approx 2$, while canonical predicts $A_2/A_1 = 4$ in the continuum limit.
• That contradicted the leading-order claims being carried from earlier derivations.
• A critic review also found that the "would-be transverse Goldstone" framing was structurally incompatible with discrete symmetry breaking.
• Three reconciliation hypotheses (α continuum extrapolation, β R22 re-derivation, γ $\Sigma_m$-Hessian convention) are under audit.

The demotion creates three follow-up paths:

• γ-path: audit the Hessian convention on $\Sigma_m$.
• β-path: re-audit the R22 cubic-equivariant derivation.
• α-path: extend finite-L extrapolation.

V5b-T' phantom: withdrawn#

The Phase 3 heuristic $\mu \approx A_{\mathrm{R3b}} \beta \lvert \partial S\rvert / \xi_0$ was withdrawn. Later numerical work suggested a mass-independent alternative, $\mu \approx 2 \beta \xi_0^2$, but W5 did not vary mass enough to distinguish the candidates.

The result: V5b-T' is not promoted. V5b-T-zero remains the safer Cat A definitional entry.

Day-5 reconciliation retractions#

Nine retractions were documented. Most were process or classification corrections; one was a substantive arithmetic correction. The point is not that W5 "lost" results. The point is that W5 made the status ledger match the actual evidence.

Open problems added#

OP-0008: σ^A K-jump inheritance non-determinism#

At K-jump merger times, pre-merger $\sigma^A(t^{*-})$ does not determine post-merger $\sigma^A(t^{*+})$ alone. The inheritance map also needs merger geometry:

• which slots merge,
• centroids,
• orientation alignment,
• post-merger relaxation trajectory.

This is not a bug in σ. It is a statement that σ must be enriched or paired with event geometry for dynamic multi-formation analysis.

OP-0009: Multi-Formation Ontological Foundations#

OP-0009 gathers seven sub-items:

The important change is that ambiguity is now named. OP-0009 prevents the same "K" from quietly meaning five different things in future derivations.

OP-0003 rider#

MO-1 remains sidestepped for single-formation σ. It reactivates if dynamic multi-formation σ or stratified Morse work becomes canonical-target work. W5 records that trigger explicitly.

Audit and reproducibility#

W5 had three verification channels, each catching a different class of problem.

The pattern for future canonical work is now:

1. pre-brainstorm,
2. Round-1 numerical sanity,
3. Round-2 structural completeness,
4. external or parallel verification,
5. reconciliation day if the week generated many working files.

Test baseline: 215 passed + 1 xfailed (216 collected), corrected from the earlier-cited "196/196" figure which was a stale snapshot per the 2026-05-04 audit.

Carry to W6+#

The W6 critical path is not "do more." It is "settle the audits that decide what can safely enter CV-1.6."

Priority queue:

Canonical impact#

W5 spans three minor canonical increments:

• CV-1.5: σ-framework support.
• CV-1.5.1: static multi-formation σ plus Commitment 16.
• CV-1.5.2: T-L1-F hard-bar / active-count bridge.

Pages affected:

• Canonical Spec — Part 5 · Results Registry
• Canonical Spec — Part 4 · Interpretation
• SCC Theorem Registry
• SCC Status (April-May 2026)
• SCC Glossary

The release-day metadata cleanup also touched theorem status headers, open-problem update history, CHANGELOG entries, and W5 weekly-summary verification notes.

Process notes#

Why T-L1-F matters#

T-L1-F is the first theorem-grade bridge between the K-field architecture and persistent topology. Before it, multi-formation theory had pieces:

• active slots,
• aggregate fields,
• persistence bars,
• local geometry conditions.

But it did not have a labeled map connecting those pieces. T-L1-F supplies the map. That is why it matters more than the raw count equality.

Why the L1 chain stayed long#

The L1-A through L1-L chain did not shrink because each step carried a different risk:

• choosing the merge level,
• bounding bridges,
• assigning slots to bars,
• preventing extra bars,
• suppressing inactive residuals,
• transferring local to global statements,
• repairing audit gaps,
• deciding the status of decay-to-cut.

Compressing that chain would have hidden assumptions. W5 chose explicit regime hypotheses instead.

Why three minor increments were acceptable#

Three minor increments in one week is noisy, but in W5 the segmentation matched the work:

• Day 1: σ support.
• Day 4: multi-formation static σ and K-status.
• Day 6: T-L1-F.

Bundling them into one release would have obscured which audit belonged to which result.

What changes going forward#

The new rule is the OP non-impact statement. Any substantial theorem candidate should explicitly say which active open problems it solves, partially addresses, clarifies, or leaves unchanged. T-L1-F and L1-M both follow this pattern.

That prevents a common failure mode: a theorem that sounds adjacent to an open problem gets read as if it solved the open problem.

Pointers#

• Full theorem registry: Canonical Spec v2.4 · Part 5
• Commitment 16: Canonical Spec Part 4 · Commitments
• OP-0008 and OP-0009: SCC Theorem Registry · Open Problems
• L1 chain provenance: SCC Status (April-May 2026)
• The previous weekly post: Perception · W4 — Critical-3 Resolved

Aligned with Perception_theory canonical CV-1.5.2 (2026-05-02). Multi-formation count theory enters canonical for the first time. The next milestone is CV-1.6, targeted for W6 close.