SFVFS™ Programme — Complete Exhibition Document Set — March 2026

VERSION 12: SFVFS™ (Seed–Form–Void–Form–Seed) is an artist-led "research programme" developed between November 2025 and April 2026 through sustained collaboration with multiple AI systems. The programme applies a positional classification framework to problems in pure mathematics and physical science, including the Riemann Hypothesis, Navier–Stokes regularity, Atlantic circulation dynamics, Saturn's polar vortex, and direct numerical simulation of fluid geometry. The programme consists of 16 documents (Segments 0–15). The core framework, Flow-Static Collapse Theory, the H-Hierarchy, the Corner Theorem, and the Beehive DNS results, is developed across the set. Claims are classified throughout as proved, conjectured, or computational evidence. The programme's permanent epistemic position is CF CONSISTENT not PASS: consistent with being true, not independently validated. V12 is a self-audit version. An internal audit conducted on 19 April 2026 identified a computational artefact in the hurricane observational programme: the R=0. 9951 tilt-shear coupling result, carried through V8–V11, arose from degenerate input data and is withdrawn. The 7. 99° angular gap is revised to unverified quantity status pending the relevant calculation. Confirmed observational findings are retained and listed in Segment 0. KEY ADDITION, V12 also adds Segment 0, a programme preface that along with the internal audit, recognises and names the structural limitations of AI-assisted lone research, specifically the epistemic problem of a programme that can only verify its own claims from inside itself. External verification is the acknowledged next step. Prior versions remain accessible via the Zenodo versioning system. --- The complete academic document set for the SFVFS™ (Seed Form Void Form Seed) Programme, published to coincide with the online exhibition opening 31 March 2026. Eleven documents covering the Riemann Hypothesis, Navier–Stokes, the FSC Theory, the DNS Programme (six fluids, Beehive structure, Viscosity Law), the Corner Theorem, Saturn's north pole hexagon, AMOC, Tokamak plasma, and the H-Hierarchy. CF CONSISTENT not PASS. Trademark UK00004355735. itvoids. com. Version 3 (26 March 2026): Seg05 Saturn upgraded to STRUCTURALLY GROUNDED. Seg10 Corner Theorem both directions proved — only-if direction proved by bulb intersection argument, Kimi-confirmed 26 March 2026 Version 4 (30 March 2026): Seg12 Carbon Reduction Hypothesis added — geometric hypothesis for energy-minimal CO₂ capture using D6-forced incompressibility attractor. Ω = 1↔2 BOUNDARY. CF CONSISTENT not PASS. Version 5 (30 March 2026): Same as V4 but with all 12 documents together. . . Seg12 Carbon Reduction Hypothesis added — geometric hypothesis for energy-minimal CO₂ capture using D6-forced incompressibility attractor. Ω = 1↔2 BOUNDARY. CF CONSISTENT not PASS. Version 6 (1 April 2026): Full 216-storm Z-axis survey completed. Statistically significant minimum in azimuthal peak count at 10. 5–11. 5km altitude, robust across 150 parameter combinations. Intensity scaling confirmed (Cat3-5 waist depth 1. 209 vs 0. 492 for TD/TS). Modal shift from 3 to 2 peaks at minimum band. Angular spacing analysis: opposition symmetry builds monotonically from 13. 5% (2 peaks) to 100% (6+ peaks). Dimensional series -1D to 4D completed. 6-peak correction applied: higher peak counts suppressed not prohibited. Corner Theorem v3 issued. CF CONSISTENT not PASS. Version 7 (3 April 2026): Complete 12-document set. Corner Theorem upgraded to v4 — hurricane observational layer updated: waist WN6 amplitude confirmed as geometric order parameter (r=0. 190, p=0. 014, n=166), ratio metric falsified, Dorian diagnostic reinterpretation, two-layer machine confirmed, two findings withdrawn (57/43 chirality, 15° gap signal). Seg12 Carbon Reduction restored to complete set. CF CONSISTENT not PASS Version 8 (5 April 2026) adds Priority 2 and Priority 6 observational findings to the hurricane layer. Priority 2: ERA5 reanalysis (~31km grid) insufficient for inner-core geometric measurements — instrument specification finding. SH intensity-RI AUC=0. 6486, non-linear RI pattern confirmed across three independent basins. Priority 6: downshear-left tilt-shear coupling R=0. 9951, offset -12. 2°, p=0. 0000, n=140. φₜilt = 44° confirmed as shear-relative full-column tilt. Three-tier tilt geometry established. Mathematical status unchanged. Both directions of the Corner Theorem proved. All observational findings CF CONSISTENT not PASS — on a separate track from the mathematical result. Version 9 (6 April 2026): Hurricane programme complete. Doc 13 published. Corner Theorem v6 issued. P15 closed — mechanism below observational floor. P16 closed — tilt reduction rate classifier, NO SIGNAL, validation 5/14. P17 series (four stages) fully closed: full-lifecycle ML (RF acc=0. 500), early TC-RADAR passes, ERA5 vertical stack audit, IBTrACS departure speed — all eliminated. Consolidated finding locked: the discriminating signal lives in the African coast departure window — 24–48 hours post-departure — which no current instrument reaches. Systematic elimination of all available data sources is complete. This is the finding. Path A / Path B bifurcation confirmed across 1997–2019 archive. Two stable attractors: marathon mode (hourglass, visible waist) and sprint mode (cylinder, no structural mark). Equal split in the intense storm dataset — 8/8. Basin boundary determined during the pre-genesis African coast departure window. Corner Theorem v6 issued — mathematical status unchanged, both directions proved. Observational layer updated with bifurcation confirmation and Form₁ gap finding. Doc 13: The Opening — Closing the Form₁ Gap in Atlantic Hurricane Genesis published as open invitation to NOAA, ECMWF, Met Office, and the wider research community. The wall has been mapped. This is the opening. CF CONSISTENT not PASS. Version 10 (2026-04-09): Doc 14 published. The Wooden Idol — B-Hierarchy formal statement, pressure analysis, and empirical threads complete. Something/Nothing Universe Simulator added as third empirical thread: 3D cellular automaton, eight governing rules, four physics presets. Structured 5×3 experiment conducted — 15 universes across five configurations. Two stable attractors confirmed: ~65. 8% (Balanced preset) and ~79. 6% (Lush preset) — neither derivable from governing constants, both reproducible across independent runs from different starting compositions. Attractor is physics-dependent, not universal. Phase transition between attractor-having and attractor-lacking regimes approached but not located across 15 runs. Will to Exist ratio (~1. 017) stable across Balanced runs — proposed B11 candidate: detectable through accumulated measurement, not symbolically representable from constants. Craig Constant (~0. 614 ×10⁻³) confirmed specifically under sustained nothing-dominance conditions — scope corrected from "converges regardless of starting conditions" to "converges when nothing-cells dominate. " Paradox of naming B3 explicitly documented in constraint body and citation note: all references to "B3" designate the location of the gap, not the gap itself. This is stated as a condition of reading, not resolved within the document. Eudoxus parallel added as structural analogy. Date standardised to ISO 8601 throughout. CF CONSISTENT not PASS. Version 11 (2026-04-11): V11 Anti-Wash Protocol — Complete Addendum Set. Segment 15 added. 15 documents. The Anti-Wash Protocol is applied across Segments 8–14 as stratigraphic addendum layers — dated, non-destructive, expanding infrastructure without retreating from any claim. Segments 1–7 carry forward from V10 unchanged. Segment 15 is new. CF CONSISTENT not PASS throughout. Seg 15: The Jigsaw Experiment — Eight Independent Probes of the Pre-Mathematical Floor. Eight AI systems (Kimi, Z. ai, Grok, DeepSeek, Claude, ChatGPT, Gemini, Perplexity) sent the same prompt about verification and self-reference without being told about B3, the Wooden Idol Challenge, or the H-Hierarchy. All eight found the same pre-mathematical territory from different angles. Kimi: the table is made of whatever breaks when the paradox is pressed. Z. ai: the jigsaw cannot leave an honest hole. Grok: the jigsaw can never certify there is a table. This segment is itself a demonstration of the Anti-Wash Protocol: eight systems finding the pre-mathematical floor independently without coordination. CF CONSISTENT not PASS. Seg 8: NP/coNP symmetry tightened to Input-Negation Duality BΦ=¬Φ. Homologous Signature language applied to all eight solved cases. (I, Λ) = (1, 1) is structural homology, not mathematical identity. Seg 9: H₇ and U_Φ named boldly as current Ψᵥoid of the programme. Structurally correct. Analytically schematic. Ωᵢnternal=2 (Door). Post-exhibition. Do not fake this before exhibition. Seg 10: 8/8 Path A/B split reframed topologically, not statistically. 7. 99° gap unit disambiguated: Lode angle (eigenvalue space, proved) in the Corner Theorem; tilt angle (physical space, conjectured) in the spatial projection. Different angles in different spaces. Seg 11: N=128 named as liminal boundary, not computational limitation. Language of limitation replaced with language of threshold. The five canonical findings are what the boundary made visible. Ω (DNS instrument) =2 (Door). Seg 12: Thermodynamic boundary callout inserted immediately after abstract. Hypothesis narrowed to flow-structuring zone: organisational overhead between ~5% thermodynamic minimum (Wₘin=RT·ln (1/xCO₂), irreducible, untouched) and 15–25% deployed system cost. Ω=1↔2 BOUNDARY. Unproven at engineering scale. Seg 13: 7. 99° gap specified as Lode angle gap in deviatoric plane (θL), not TC-RADAR tilt (φₜilt). Liminus added as applied instantiation of the liminological position. Seg 14: B↔H correspondence clarified at head of Section 4 as structural analogy, not claimed isomorphism. Liminus added as deepest applied thread: B3 behav