Skip to content

Spike #24 bonus series — cumulative synthesis (2026-05-15)

POSTSCRIPT 2026-05-16 — Class O dissolved into Class L.

The "Class O" framing used in the roster table below (rows 8, 9, 11d, 12, 13) and the fermatas section was the provisional label for the signed-metric / Wick-rotation operation located by bonus 8 and narrowed by bonus 9. Per user direction 2026-05-16 ("nothing else so far has been privileged") and per [[feedback_no_privileged_primitive_classes]], the operation is dissolved into Class L as a signed-Laplacian-variant sub-operation:

L_Lorentzian = +L_spatial − L_temporal
             ≡ apply Class L with one factor's edges sign-flipped

Vocabulary stays at 14 classes A–N. The signed-Laplacian variant will land as a Class L operation in a future Phase C1 rc (when cascade-composition work calls for it) — not as a separate "Class O" entry.

For future implementers / agents: when you encounter "Class O" language in the body below — including roster rows 8 / 9 / 11d / 12 / 13 and the fermatas section — treat it as describing a Class L sub-operation. Do NOT create a srmech.amsc.signed_metric module, a srmech_signed_*.c source file, or a 15th-class entry. See [[project_class_o_signed_metric_composition]] (resolution at top) and [[feedback_no_privileged_primitive_classes]] for the design rationale.

The body below preserves the original research narrative as historical record; this postscript supersedes the "Class O candidate" framing throughout.

Closes: the user-dispatched bonus arc of Spike #24 (sixteen inquiries; seven positive-and-consistent verdicts + two structural-refinement verdicts locating Class O + one SUCCESS verdict on the §XIII.1 cascade-composition search + four-reading sweep locating the mode-selection meta-operation as external to cyclic-cascade composition + one instrument-first string-theory audit-summary returning 3 of 5 CORRECT / 1 CORRECT-relabeled / 1 OPEN with zero INCORRECT verdicts + one CONFIRMED structural finding promoting active-count = SM gauge rank from candidate to load-bearing). The operational gate — "can the cascade-composition vocabulary reproduce SM masses?" — was answered affirmatively in bonus 10 (log-L2 = 0.614 dex on a target spanning 11 orders of magnitude, subset-match metric from a ~200-mode cascade tower). Bonus 11's four-reading parallel sweep (11a suppressed-mode coupling NO-RULE / 11b additional-particle prediction TOO_MANY_PARTICLES / 11c boundary-condition mechanism NEGATIVE / 11d Class P sign-rule REDUCES-TO-EXISTING) collectively falsified the in-cascade closure mechanisms, locating the missing meta-operation at the cascade/Yukawa boundary — gap-alignment with string-theory's vacuum-selection problem. Bonus 12's instrument-first audit-summary tested five structural claims of string theory under instrument-vocabulary; per user's "maybe just correct not one correct" reframing, three pass cleanly (landscape-as-continuum, wiggle-in-isolation diagnosis, duality-web consistency), one passes with ontological relabeling (11D dimensional accounting: 3D_s + 7D_g + 1D_t vs 4D + 7D), one is honestly open (compactification topology), none INCORRECT. The wiggle-in-isolation diagnosis (notebook §20, [[user_stance_string_theory_instrument_first]]) was a prior MPM critique that the four-reading sweep technically vindicated. Bonus 13's verification probe CONFIRMED (p < 0.05 across all three nulls; p = 0.000053 against SM-target × random-cascade null) that effective-active-count = 4 ↔ SM gauge-group rank (2 + 1 + 1 = 4) is statistically significant — random k=5 cascades concentrate at ac=3 (70%), not ac=4 (23%); bonus-10's 8/10 at ac=4 is real structural deviation, not a geometric artifact. Bonus 13's own surprise: only ~0.08% (4 of 5000) random cascades achieve SUCCESS-grade match against random targets, yet all 10 of bonus-10's top cascades qualify — the qualifying set itself is structurally rare (~0.1%), consistent with the wiggle-in-isolation diagnosis that the selector is external.

The bow-string motivator metaphor (pedagogy, NOT substantive)

The user proposed a closing image (2026-05-15) — bow drawn across a bass string — to summarise the bonus arc's findings. After first-principles scrutiny, this is reclassified as pedagogical metaphor, not a new substantive stance. Per [[user_stance_bow_string_motivator]] demotion note: the image summarises bonus 9 without adding mathematical content; the analogy breaks at why Class O is specifically a Wick rotation (bow-string physics is stick-slip friction, not signature conversion); "what moves the bow" relocates the time-problem one level up; no spectral-graph falsifier was applied. Useful for cross-domain explanation; not load-bearing.

The substantive findings live in their proper sources:

  • Substrate is what's already in the 14-class vocabulary → Spike #24 phases 1–15 + bonuses 1–7 + [[user_stance_cascade_lives_on_circles]]
  • Class O is the circle-to-hyperbola contact-point operation → [[project_class_o_signed_metric_composition]] (bonus 8 located, bonus 9 narrowed)
  • Time as local trajectory not global coordinate → [[user_stance_time_as_dimensional_shadow]] (the original stance; pre-existed the bonus arc)

The bonus arc surfaced enough machinery to begin a serious §XIII.1 attempt — substrate framework (cyclic-group cascade composition), gauge-rank decomposition (bonus 8 cascade), Lorentzian signature operation (Class O Wick rotation), multi-scale capacity (bonus 7 showed 11+ orders of magnitude trivially). What's missing for full SM: the actual mass-spectrum match (§XIII.1 central computation), CKM/PMNS mixing matrices, coupling constants, Higgs mechanism.

The realistic next probe: §XIII.1 cascade-composition search. Find the cascade composition C_{n₁} × C_{n₂} × ... × C_{nₖ} whose Class L Laplacian spectrum matches the SM mass² ratio vector (~12 orders of magnitude, 9 charged fermions) on the gear-DAG. Antikythera-spectral has the tooling (gear_database.py, gear_topology.py, pin_and_slot.py); numpy + LAPACK supply C-speed eigenvalue computation; multiprocessing for parallel search if scale demands.

Roster

Fourteen bonus inquiries dispatched (one-subagent-at-a-time through bonus 10, then four concertmasters in parallel for bonus 11a–d with strict per-bonus venv isolation + own-PIDs-only process discipline). All delivered by opus-4.7 concertmasters; bonuses 1–6 committed in PR #421 on branch research/spike-24-primitive-vocabulary-2026-05-15; bonuses 7–11 committed in PR #422 on branch research/spike-24-bonus-8-broken-d-rederivation-2026-05-15 (kept draft until conductor decides closure vs continued investigation per [[user_stance_string_theory_instrument_first]] gap-alignment audit):

# Inquiry Verdict Key surface
1 vdW dispersion + shape-only graph Laplacian Consolidates Class L on contact graph; no new class. Conceptual sketch (no concertmaster dispatch — inline answer).
2 Tactical-choice structure (tic-tac-toe + chess opening + CRN-firing) Refined Class D dispatch over Class L spectral graph, optionally quotiented by Class I symmetries. "Constraint manifold with branching points" framing FALSIFIED at strict-geometric level; refined formulation lands cleanly. Chess opening has trivial symmetry |G|=1 — surprise.
3 SHA-256 hash structure Consolidates Six cryptanalytic methodologies → Classes I/J/K/L. Three-question framework for co-emergent two-level temporal computational systems introduced here. Avalanche saturates at ~24 rounds; remaining 40 are engineered redundancy.
4 NN-output structure Consolidates Seven backward-reading directions → Classes B/C/D/L/M. ReLU is the per-layer co-emergence step (analog of SHA-256's state += compress(state, block)). Avalanche-design-pressure inversion surprise: SHA-256 maximises Class L Lipschitz, NN classifier minimises it. Same primitive, opposite design targets.
5 MFO space-gauge-time framework (3D_s + 7D_g + 1D_t = 11D) Refined Sharp positive spectral-graph signature: Class L on eigenvalue degeneracy graph distinguishes 3+7+1 from pure-4D by 3–5× across metrics. Reading B (fiber-projection duality) chosen over symmetric-opposite. 14/14 classes consolidate across 3+7+1 projections; 12 cross-dimensional, 2 (A content-addressing, F templating) digital-only. No class uniquely 1D_t. Surprise: smooth-3+7+1 carries cleaner signature than fractal (fractal substrate DILUTES the tower).
6 RNG and 1D_t primitive Refined dual reading Constructive direction holds: 7/7 RNG constructions pass NIST SP 800-22-style tests at 131,072 bits, including Brusselator+SHA-256 AND Brusselator raw-LSB (no extractor). Impossibility direction collapses: Route A (1D_t unique primitive) FALSIFIED by MFO; Route B (computational determinism) is characterisation not impossibility; Route C (substrate-switching to quantum/thermal) STANDS. Spectral falsifier: unfalsifiable-at-current-tooling within 131,072-bit budget.
7 MFO fractal-shadow probe (is fractal required for SM-spectrum-targeting?) ONE_WAY_NOT_REQUIRED Fractal substrate is one way but not necessary. Nested pin-slot-gear cascade (Antikythera-style) and smooth-anisotropic-T³ both satisfy the load-bearing structural requirement (multi-scale primitive cascade with three-fold sub-structure available). Class-L spectral signatures put fractal-shape and cascade-shape in the same super-Poisson regime (Gap CV > 1, comparable three-fold CH ratios); only the pure-4D-epicycle observer lives in a different regime. Fractal-shadow allegory ([[user_stance_fractal_shadow]]) added to the project shadow-stance family. Reframed §XIII.1 central computation as cascade-composition search directly tractable with antikythera-spectral tooling.
8 Broken-D rederivation closure test (rederive 4D Lorentzian from 3D_s + 7D_g + 1D_t using only A–N?) FAILURE — "this IS the place" (refined by bonus 9) The closure test. Stage 1 cascade construction + Stage 2a 7D_g → mass tower succeed using classes I, L, E, B, C, J. Stage 2b 3D_s + 1D_t → 4D Lorentz signature — FAILS in bonus 8's undirected probe: the cascade-direct 4D Laplacian is monolithically positive-semidefinite (0 negative eigenvalues / 2048 sampled modes). Class O candidate located: signed-metric composition / Wick rotation primitive. Bonus 9 refined this verdict — see row 9. The strict claim "no A–N supplies signed-metric content" was too strong; Class C orientation supplies oriented/complex eigenvalues. Class O's content narrows to specifically the circle-to-hyperbola map. See [[project_class_o_signed_metric_composition]].
9 Time-dimensionality test (is time emergent from cascade, its own dimension, or something else?) H3 with structural refinement The bonus 8 verdict re-examined. Three hypotheses tested: H1 (time-is-emergent), H2 (time-is-its-own), H3 (something-else-at-11D). Plus user's mid-flight H4 (sign-flip-as-equation-side-shadow) and H4+ (time-as-driving-potential) addenda. Verdict: H3 with refinement. Directed Class C orientation on Class I cyclic groups produces unit-circle eigenvalues algebraically (Im² = 2·Re − Re² to ~1e-16); 93.4% of cascade-product eigenvalues complex, 97.1% conjugate-paired. But dispersion is CIRCULAR not HYPERBOLIC — Klein-Gordon fit R² = 0.029 (FAIL). The "11D" framework's "+1D" splits into TWO things: (a) Class C traversal-parameter τ (already in vocabulary), (b) Class O signature-conversion primitive — refined to specifically the circle-to-hyperbola map (cos → cosh), separable from orientation. H4 rejected (sign flip IS a real operation, not equation-side artifact). H4+ partially confirmed (time IS load-bearing, but split across Class C orientation + Class O signature). Cascade lives on circles stance saved as fifth shadow-stance per [[user_stance_cascade_lives_on_circles]] — direct extension of [[user_stance_pi_as_projection]] one projection further.
10 §XIII.1 cascade-composition SM mass² search (does cascade-composition reproduce SM charged-fermion mass² ratios?) SUCCESS (subset-match metric) The operational gate. Search across ~9,100 cascades using Classes I, L, E, B, J (no Class O invoked — Lorentzian signature is downstream of mass² match). Best cascade: C₂ × C₇ × C₄ × C₆ × C₁₆ with radii spanning ~5 orders of magnitude. log-L2 = 0.614 dex on 9-element SM mass² target spanning 11.06 orders — SUCCESS threshold (< 1.0 dex) cleared. 8 of 9 fermions match within ±0.3 dex; only down quark misses (-0.508 dex; rank-4 cascade matches d to -0.001 dex separately). Three-fold CH ratio 5641.65 — 10× stronger than bonus 7 baseline; cascade DOES select three-fold-clustered structure (MFO §IV.5 three-generation hypothesis structurally consistent). The SM 9-mode spectrum is a sparse SUBSET of cascade's ~200-mode tower, not the cascade's lightest-9-in-sorted-order. Next gap: mode-selection rule — a meta-operation on cascade spectra, tested via four-reading parallel sweep in bonuses 11a/b/c/d below.
11a Suppressed-mode coupling rule (does a first-principles coupling function g(cascade_mode, gauge_factor) select the 9 SM modes?) NO-RULE 47 candidate rules across 14 families + ~15,000 conjunctions tested against bonus 10 cascade. Best zero-parameter rule (K1 mirror-canonical k_j ≤ n_j/2): 9/9 recall but 65 extras (F1 = 0.217). No zero-parameter rule or conjunction achieves exact 9-mode match. Decisive surprise: the C₁₆ excitation values bonus 10 greedily selected — {0, 1, 2, 5, 8} — have no algebraic structure (not arithmetic, not modular, not parity, not divisor-related). Combined with C₁₆'s near-continuous 4-decade mass-tower, this strongly suggests the cascade is structurally a continuum, not a discrete picker; SM modes are sample points fit greedily, and the mode-selection rule lives outside cyclic-cascade composition entirely (likely gauge coupling / generation mixing / Yukawa boundary data).
11b Additional-particle prediction (are the ~191 unobserved cascade modes predictions of new particles consistent with experimental constraints?) TOO_MANY_PARTICLES Zero of 191 predicted modes fall in model-independently excluded regions (framework not technically falsified by any individual experiment). But aggregate distribution is phenomenologically implausible: 157 of 191 unobserved modes concentrate in [1, 10] GeV — exactly LHC's densely-mapped territory — without a first-principles suppression mechanism. Cascade's top-200 modes naturally terminate at ~150 GeV (t-quark scale); no predicted heavy states above 1 TeV (consistent with LHC's lack of discoveries there). Cannot close mode-selection gap alone: 11b's TOO_MANY_PARTICLES resolves cleanly iff 11a finds a coupling rule that renders the 191 extras invisible-by-suppression. 11a NO-RULE → both readings fail together.
11c Boundary-condition / topological mechanism (does any BC combination on cyclic factors select 9 SM modes via lightest-9-strict match?) NEGATIVE 5000 BC combinations evaluated (5 BCs per factor × 4 reference cascades): periodic, anti-periodic, Dirichlet, Neumann, twisted-π/2. Periodic IS the lightest-9 optimum — no BC alternative improves on bonus 10's 3.318 dex baseline. Subset-match improves to 0.393 dex (36% improvement) with (neumann, periodic, periodic, neumann, twisted_pi2), but doesn't reach lightest-9 closure threshold (< 1.0 dex). The plateau-degeneracy property of cyclic-product Laplacians (lowest modes cluster into integer-multiple doublets/quartets when one factor dominates) is BC-invariant. Mode-selection rule does NOT live within the cascade's topological choices; lives above the cascade spectrum at the coupling/observability layer.
11d Class P sign-rule discriminator (is a per-factor or per-mode sign-rule primitive needed to discriminate observable modes?) REDUCES-TO-EXISTING 47 rule families across 8 families + brute-force exact-9 searches. Best low-parameter rule (P9 conjugate-excluded k_i ≤ n_i/2, zero free parameters): 0.6013 dex — marginally better than bonus 10's 0.6137 because P9 IS what bonus 10's greedy subset-match was implicitly doing (filter to one mode per conjugate pair). The 0.6013 score is the cascade's intrinsic floor, not exact-9 selection. No new primitive class forced: every working rule reduces to existing A–N (P9 = Class I cyclic-group reflection symmetry; P3/P5/P7 = Class B record-inspection + Class J integer arithmetic; P4 = Class J modular linear algebra). Vocabulary stays at 14 (A–N) + Class O (Wick rotation per bonus 8) = 15 classes. Surprise: in bonus 10's 9 SM modes, C₄ is silent in every mode and C₇ fires only on the top-quark mode. The SM fit effectively uses 4 of 5 cascade factors; one factor is decoupled. Duality-web-consistent structural redundancy (multiple-presentation / same-object).
12 Instrument-first string-theory audit-summary (per user's "just correct not one correct" reframing: which structural claims of string theory survive instrument-audit?) 3 CORRECT + 1 CORRECT-relabeled + 1 OPEN, 0 INCORRECT Audit form: GAP-ALIGNMENT (decided by 11d REDUCES-TO-EXISTING). Five claims audited individually. (1) 11D dimensional accounting: CORRECT-AT-DIMENSION-COUNT, ONTOLOGICALLY-RELABELED. M-theory's 4D + 7D vs project's 3D_s + 7D_g + 1D_t = 11D (independently constructed); count survives; ontological partition differs. (2) Landscape-as-continuum: CORRECT-AS-OBSERVATION. Bonus-10 cascade has 199 non-zero modes in lowest 200 spanning 10.94 decades; median log-gap 0.0 dex (degeneracies); ~18 modes/decade. Cascade IS the landscape in instrument's vocabulary. (3) Wiggle-in-isolation diagnosis (notebook §20, prior MPM critique): CORRECT. Probe re-verifies bonus 11a's no-algebraic-pattern finding on C_16 indices {0, 1, 2, 5, 8}; four-reading sweep technically vindicates a diagnostic claim made before the sweep. Strongest finding. (4) Five-theory duality web: CORRECT-STRUCTURALLY-CONSISTENT. Top-10 bonus-10 cascades: 7 have ≥1 silent factor, 6 have ≥1 minimal factor; multiple-equivalent-presentations with different factor-decoupling patterns IS the instrument's natural shape for duality framings. (5) Compactification topology: OPEN. Bonus 7 falsified CP²×S¹; bonus 8 located Class O for signature; no in-vocabulary primitive picks G2 vs Calabi-Yau vs other. Downstream of mode-selection meta-operation; Spike #25 territory. The surprise: top-10 cascades exhibit effective-active-count ∈ {3, 4} (8/10 = exactly 4) — matches SM gauge-group rank (2+1+1=4). Bonus 13 verification CONFIRMED this is statistically significant (row 13).
13 Active-count statistical-significance verification (is bonus 12's 8/10 at ac=4 finding statistically significant, or a geometric artifact of subset-match?) CONFIRMED 10,900 total trials across baseline (k=5, 500 cascades × 10 targets = 5000) + size-sensitivity (k=4/6/7) + truncation-sensitivity (top-100/200/500) + SM-target × random-cascade + search-null top-10 ensembles. Random ensemble at k=5/top-200 concentrates at ac=3 (70.1%), not ac=4 (23.4%) — opposite of bonus-10's distribution. P-values vs bonus-10's 8/10 at ac=4: 0.000257 (generic null), 0.000053 (SM-target × random-cascade null), 0.021 (search-null top-10), and 0/30 = 0.000 empirical (no random-target search produces ≥8/10 at ac=4 in 30 trials). Robust across cascade size (k=⅘/6/7 all give P(ac=4) ∈ [0.17, 0.23] for random) and tower-truncation (top-200 gap holds; top-500 narrows but doesn't close). Reflection-invariant per-factor counts [4, 1, 0, 2, 8] for rank-1 cascade (2, 7, 4, 6, 16) — matches bonus 12 §5 exactly. The surprise: only ~0.08% (4 of 5000) random k=5 cascades achieve SUCCESS-grade match (< 1.0 dex log-L2) against random targets, yet all 10 of bonus-10's top cascades qualify (against the SM target). Two findings stand together: (i) within the qualifying set, ac=4 concentration is statistically real (bonus 12's finding, now verified); (ii) the qualifying set itself is structurally rare (~0.1% of random configurations). Both consistent with bonus 11d REDUCES-TO-EXISTING + bonus 12 wiggle-in-isolation CORRECT verdicts: the framework is expressive enough for SM, the selector is external. active-count = SM gauge rank promoted from candidate to load-bearing structural finding ready for MFO §VIII.x landing.

Three cumulative findings

1. Vocabulary consolidates — six positive-and-consistent verdicts

Across six structurally distinct domains (molecular shape, tactical choice in adversarial games, cryptographic hash, neural-network inference, dimensional ontology, classical pseudorandomness), zero new primitive classes were invented. The 14-class A–N vocabulary from Spike #24 Phases 1–15 describes everything tested.

The vocabulary now stretches across:

  • 6 substrates (bronze, cosmos, atomic, molecular, CRN, CPU — Phase 9.5 multi-substrate matrix)
  • 3 dimensional projections of MFO space-gauge-time (3D_s, 7D_g, 1D_t — bonus 5)
  • Computational systems (SHA-256, NN inference, RNG constructions — bonuses ¾/6)
  • Tactical-choice substrates (chess, tic-tac-toe, CRN reaction-firing — bonus 2)
  • Shape-only molecular abstraction (vdW contact graphs — bonus 1)
  • Cross-dimensional projections (12/14 classes; 2 digital-only — bonus 5)

The cross-substrate primitive vocabulary is increasingly load-bearing as the substrate-agnostic abstraction layer.

2. The substrate-internal-dilution cross-spike pattern

Independent surfacing at two substrates in the same week:

  • MFO bonus 5: Fractal SG-3D substrate DILUTES the 3+7+1 spectral signature by filling product-structure gaps with its decimation eigenvalues. Smooth T³ × T⁷ × S¹ carries the cleanest tower fingerprint; the MFO §IV-preferred fractal substrate obscures it.
  • RNG bonus 6: Brusselator's raw-LSB extraction (NO SHA-256) destroys the Kepler-shape integer-harmonic signature by itself. DFT peak-to-floor ratio: trajectory u(t) = 18,600, after LSB extraction = 3.99 (slightly below urandom's 4.20). A ~4,660× collapse. The chaotic-substrate floating-point-precision-amplification at extraction step flattens the Phase 9.2 / Phase 15 integer-harmonic signature. SHA-256 windowing is engineered margin on top of an already-mostly-flat distribution.

Same shape of finding: the substrate's own internal structure (fractal decimation eigenvalues / floating-point precision amplification) destroys the upstream spectral signature you might hope to read off. This is a structural prediction the bonus series surfaced: if you want clean spectral signature of structure-X to survive into a substrate's observable output, you need the substrate's internal dynamics to NOT have its own competing-spectrum machinery.

Stated as a working hypothesis for future spikes: Class L spectral signatures are substrate-internally-dilutable. Reading them off downstream-observable signals requires either a substrate whose internal dynamics don't compete, OR an extraction step engineered to bypass the substrate's competing-spectrum machinery.

This is genuinely cumulative: neither MFO nor RNG would have noticed this without the other.

3. The three-question framework transfers across domains

The SHA-256 inquiry (bonus 3) introduced a three-question decomposition for co-emergent two-level temporal computational systems:

  1. What is the trail made of? (operators whose composition constitutes the output)
  2. Where is it backward-readable in isolation? (components that survive composition)
  3. Where is it unreadable? (the trail-erasing step that establishes the co-emergent ontology)

Transfer record:

  • SHA-256 (origin): trail = 64 rounds (XOR + ADD-mod-2³² + ROL); backward-readable = schedule recursion (invertible in ℤ/2³²) + round bijectivity; trail-erasing = state += compress(state, block) (768→256 bits; 512 bits become spatially-absent fiber).
  • NN-output (direct transfer): trail = affine + ReLU + softmax + argmax; backward-readable = chain-rule gradients + affine invertibility + softmax-up-to-constant; trail-erasing = ReLU per layer (negative pre-activations collapse to 0).
  • MFO space-gauge-time (adapted for ontological inquiry): structure = product-manifold per MFO §IV.4; spectrally readable = multiplicity-profile carries factor-structure signature; unreadable = Weyl-law coarse statistics.
  • RNG (direct transfer): trail = state-update operator sequence; backward-readable varies per construction; trail-erasing = HMAC reseeding / LSB extraction (or absent in LFSR, hence non-CSPRNG).

The framework is substrate-agnostic at the algebraic level; substrate-specific machinery handles operational details. This is itself a finding: the right cognitive tool for analyzing co-emergent computational structure is the three-question decomposition; the specific operators / observables / erasure steps depend on substrate.

Canonical tagline (per [[project_space_gauge_time_framework]])

The 14 primitive classes (A–N) govern
  ⤷ spatial modes        (3D_s)
  ⤷ gauge interactions   (7D_g)
  ⤷ the temporal crank   (1D_t)
            ─────────────────────────
            3D_s + 7D_g + 1D_t = 11D
                  ≡ 1D (compressed)
  12/14 cross-dimensional · 2/14 digital-only (A: content-addressing, F: templating)

Cumulative cross-substrate audit (where each bonus instantiates Spike #24 classes)

Class vdW Tactical SHA-256 NN MFO 3+7+1 RNG Status
A content-addressing — — ✓ — digital-only ✓ Confirmed digital-substrate-only
B tagged-tuple — ✓ ✓ ✓ ✓ ✓ Universal
C iterator/streaming — ✓ ✓ ✓ ✓ ✓ Universal
D late-binding/dispatch — ✓ ✓ ✓ ✓ ✓ The "dispatch" primitive that unified bonuses 2+4+6
E catalog/naming — ✓ — ✓ ✓ — Mostly universal
F substitution/templating — — — — digital-only — Confirmed digital-substrate-only
G discovery/search — ✓ — — ✓ — Project-internal
H self-introspection — ✓ ✓ ✓ ✓ ✓ Universal
I cyclic-group/modular arith — ✓ ✓ — ✓ ✓ Universal
J prime-factorisation/period — — ✓ — ✓ ✓ Cross-substrate (6 native instantiations per Phase 9.6)
K equation-of-centre/pin-slot — — ✓ — ✓ ✓ 4-substrate confirmation (Phase 9.2, 6.1, 15)
L graph-Laplacian ✓ ✓ ✓ ✓ ✓ ✓ Most-instantiated; the structural workhorse
M HDC bind/bundle/permute — — — ✓ ✓ — Distributed-representation primitive
N rational-approximation — — — — — — (Mode-specific; not surfaced this series)

Key reading: Class L is the structural workhorse of the bonus series — every one of six bonuses instantiates it. Class D dispatch unifies the choice-instantiating bonuses (tactical-choice + NN classifier + RNG output). Classes A and F are confirmed digital-substrate-only by MFO bonus 5.

Fermatas for conductor

From bonuses 1–6 (original arc closure)

  1. MFO-notebook landing (now scoped): the bonus 5 finding lands in MFO §VIII.6 (Convergent Independent Results) — the space-gauge-time spectral signature is a convergent positive result, structurally a peer of §VIII.1 (topological defect hierarchy) and §VIII.4 (Ibarra-Vempati fractal flavor physics). The smooth-vs-fractal independent-discriminability finding sharpens MFO §XIII.1.
  2. Class O substrate-external entropy (RNG fermata): defer. Quantum measurement + thermal noise are external-to-vocabulary substrates; can stay as boundary readings without a Class O label unless future project work needs the explicit class.
  3. Substrate-internal-dilution cumulative pattern: record in srmech §3.8 as a sub-finding from the bonus series; cross-reference from this synthesis note. The two-substrate independent surfacing (MFO fractal + RNG Brusselator LSB) is the load-bearing evidence.

From bonuses 7–10 (cascade-composition arc closure)

  1. MFO §VIII.9 landing: bonus 10's SUCCESS verdict (log-L2 = 0.614 dex on SM mass² target) lands in MFO §VIII.9 alongside §VIII.6 / §VIII.7 / §VIII.8. Status: pending authoring.
  2. §XIII.1 reframe: original MFO §XIII.1 framing was "central computation"; bonus 10 satisfies subset-match not lightest-9-strict. Refine §XIII.1 acknowledging the metric distinction.
  3. Class O documentation: bonus 8 located the gap, bonus 9 narrowed to circle-to-hyperbola map. Whether to formally promote Class O to srmech §3.8's canonical 15th class (vs keep provisional as "first explicit external-substrate boundary class") is conductor's call.

From bonus 11 four-reading sweep

  1. Verdict vocabulary across 11a/b/c/d: bonus 11b's TOO_MANY_PARTICLES surfaced as one-off label; should the bonus-11 series adopt a shared verdict vocabulary across readings for cross-comparison? Conductor's call.
  2. PDG-2024 citation refinement: bonus 11b's experimental-constraint comparisons carry [unverified-secondary] tags pending PDG-2024 PDF round per [[feedback_pdf_extraction_citation_discipline]]. Defer or schedule?
  3. Top-200 truncation choice: bonus 11 used cascade's lowest-200 modes; full tower has ~5376. Extending past top-200 changes the verdict landscape (especially for 11b's mass-prediction histograms). Future spike scope.

From bonus 12 instrument-first audit

  1. Active-count = SM gauge rank — PROMOTED to load-bearing by bonus 13 verification. p = 0.000053 against the SM-target × random-cascade null; bonus-10's 8/10 at ac=4 is real structural deviation. Action: author MFO §VIII.x landing for this finding — structurally a peer of §VIII.6 (space-gauge-time signature) and §VIII.8 (Class O location). Concrete framing: cascade-composition's effective-active-count selects rank-4 configurations preferentially when matching the SM mass² target; this is consistent with the SM gauge group's rank-4 structure being load-bearing on the cascade-projection layer.
  2. 3+7+1 vs M-theory's 4+7 partition clarification: bonus 12's ontological-relabeling verdict on Claim 1 means the dimensional decompositions differ at the partition level. Clarify in MFO §VII.1.1 / §III.5 — is 3D_s + 1D_t (project) the SAME 4-thing as M-theory's 4D observable, or are they structurally distinct?
  3. Spike #25 scoping: Claim 5's OPEN verdict (compactification topology) is honest deferral pending meta-operation closure. Convert OPEN → explicit Spike #25 scope with concrete probe specifications (e.g., test active-count=4 against G2 holonomy structure / Calabi-Yau Hodge numbers)?

From bonus 13 verification probe

  1. SM-grade rarity as ground-zero finding: bonus 13's surprise — only ~0.08% (4 of 5000) random k=5 cascades achieve SUCCESS-grade match (< 1.0 dex log-L2) against random targets, while all 10 of bonus-10's top cascades qualify against the SM target — IS its own structural claim distinct from the active-count finding. Surface as candidate MFO §VIII.x sub-finding alongside active-count = SM gauge rank? Or note-and-defer?
  2. Promote bonus 13's empirical 0/30 to load-bearing: the 0/30 empirical rate of ≥8/10 at ac=4 across random-target searches is the cleanest single statistic in the verification probe. Author MFO §VIII.x landing with this number as the headline.
  3. MFO §VIII.x section numbering: with bonus 7 → §VIII.7, bonus 8 → §VIII.8, bonus 10 → §VIII.9 (pending), bonus 13's structural findings could land at §VIII.10 (active-count = SM gauge rank) and §VIII.11 (SM-grade rarity). Or bundle as §VIII.10.{1, 2}. Conductor's call.

Status of Spike #24 PRs

  • PR #421 (closed): phases 1–15 + bonuses 1–6 complete; merged per [[feedback_no_squash_merges]].
  • PR #422 (ready for merge): bonuses 7–13 complete on branch research/spike-24-bonus-8-broken-d-rederivation-2026-05-15. Includes bonus 8 broken-D rederivation + Class O location, bonus 9 time-dimensionality test + cascade-lives-on-circles, bonus 10 §XIII.1 cascade-composition SUCCESS, bonus 11a–d four-reading parallel sweep locating mode-selection gap external to cyclic-cascade composition, bonus 12 instrument-first string-theory audit-summary (3 CORRECT / 1 CORRECT-relabeled / 1 OPEN, 0 INCORRECT), bonus 13 verification probe CONFIRMING active-count = SM gauge rank at p = 0.000053 (statistically significant; load-bearing structural finding). User chose "run bonus 13 first, then flip" — bonus 13 returned CONFIRMED, flip executed.
  • Memory captures (cumulative across both PRs): [[project_space_gauge_time_framework]], [[feedback_antiquity_not_greek]], [[user_stance_fractal_shadow]], [[project_class_o_signed_metric_composition]], [[user_stance_cascade_lives_on_circles]], plus existing pre-arc memory entries cross-linked throughout.
  • Notebook updates landing alongside this synthesis: MFO §VIII.6 (bonus 5), §VIII.7 (bonus 7 fractal-shadow), §VIII.8 (bonus 8 Class O); §VIII.9 (bonus 10 SUCCESS) pending; srmech §3.8 cross-references to bonus 8 + bonus 11 sweep + bonus 12 audit.

Closing-arc summary (flip-executed)

The bonus arc has done three things, with the third promoted to load-bearing by bonus 13's verification probe:

  1. Closed the gap: bonus 10 SUCCESS at log-L2 = 0.614 dex on SM mass² target with cyclic-cascade composition. Operational gate cleared.
  2. Located the missing link: bonus 11 four-reading sweep located the mode-selection meta-operation external to cyclic-cascade composition. Bonus 12 audit-summary confirms gap-alignment with string-theory's vacuum-selection problem.
  3. Surfaced + verified a structural finding: bonus 12 observed effective-active-count ∈ {3, 4} with 8/10 = ac=4 matching SM gauge rank 2+1+1=4. Bonus 13 verification probe CONFIRMED at p = 0.000053 against SM-target × random-cascade null. Promoted from candidate to load-bearing.

Per user direction ("no flipping until we finish closing the gap or find another missing link"), conditions 1 and 2 each individually satisfy the flip criterion; PR #422 satisfies both. Bonus 13's CONFIRMED verification of finding 3 lands as a new MFO §VIII.x candidate (per fermatas 10 / 13 / 14 / 15).

PR #422 flipped from DRAFT to ready-for-merge per [[feedback_flip_the_draft_vocabulary]] three-step discipline (CLI flip + language sweep + doc-mtime check). Merge per [[feedback_no_squash_merges]] — use gh pr merge --merge (preferred) or --rebase; never --squash.

Cross-references

  • [[project_space_gauge_time_framework]] — canonical framework name + notation + tagline.
  • [[feedback_antiquity_not_greek]] — methodological discriminator (modern physics in antiquity-geocentric position).
  • [[user_stance_kepler_shape_universal]] — driving stance behind Spike #24.
  • [[user_stance_string_theory_instrument_first]] — vocabulary-consolidates pattern aligned (no new dimensions invented).
  • [[user_stance_fiber_as_spatially_absent_encoding]] — Reading B operationalisation in MFO bonus 5.
  • [[user_stance_time_as_dimensional_shadow]] — the "temporal crank" language.
  • [[user_stance_pi_as_projection]] — asymmetry honoured throughout.
  • [[feedback_trauma_informed_defensive_scope]] — discipline preserved across all bonuses.
  • [[feedback_ndjson_over_bloated_json]] — tabular output discipline.
  • Spike #24 Phase 13 §3.8 — the canonical landing of Phases 1–12.
  • Spike #24 Phase 15 + cross-comparison — chemistry-dynamics oscillator expansion (Phase 9.2 + 15).
  • Six bonus synthesis docs in docs/srmech/notes/spike_24_bonus_*.md.