# Halcyon

> The first Gi_System. A geometric apparatus for engineered modification of local
> inertial response. A small spherical cage of ninety Josephson junctions arranged
> on the geometry of a buckyball (truncated icosahedron), realising an SU(2) gauge
> field on a 60-vertex / 90-edge / 32-face cell complex of Euler characteristic 2.
> The framework predicts that the topological-sector-dependent ground state of the
> gauge field couples to inertial mass; the experiment is the m_eff/m_0 curve across
> programmed sectors Q.

Author: Bee Davis (Davis Geometric). Patent pending. Page schema 2026-06.

## Landing surfaces

- [Halcyon landing page (HTML)](https://davisgeometric.com/halcyon): interactive view with hero, history timeline, topological-sector explainer, live cage simulator, math chain, validation grid, and Gi_Systems family.
- [Halcyon landing page (markdown)](https://davisgeometric.com/halcyon/llms-full.txt): same content as plain text (recommended for AI ingestion).
- [Cage simulator (embedded demo)](https://davisgeometric.com/halcyon/cage_preview): standalone Three.js + WebGL view of the buckyball substrate with SU(2) trajectory replay, scenario buttons (Quench Up / Baseline / Quench Down), per-edge phase / kinetic / both color modes, an Advanced drawer with a Generate-Validation-Report button.

## A thread, pulled two ways (what the experiment is actually measuring)

The textbook two-thread inertia demo separates gravitational loading from dynamic response. A mass hangs from a thread tied to the ceiling; a second thread hangs from the bottom of the mass to a hand. Pull slowly: the upper thread breaks (it bears mg + F_pull). Pull fast: the lower thread breaks (the mass cannot get the news to the ceiling in time, so the lower thread bears F_pull alone). Same weight, same gravity, different breaking thread — the difference is the pull rate compared to the bulk's intrinsic stress-redistribution time.

The relevant dimensionless number is the Deborah number D = tau / t_load = r * tau / F_*, with r = dF/dt, F_* the breaking-force scale, and tau the inertial relaxation time of the bulk. D << 1 → quasi-static (upper breaks); D >> 1 → impulsive (lower breaks). In the idealised elastic model the inertial crossover sits at omega_c = sqrt(K/mu), tau_mu = sqrt(mu/K), so mu = K * tau_mu^2. For a real apparatus the transfer function is chi_Q(omega) = 1 / (K_Q + i * c_Q * omega - mu_Q * omega^2). For a measured normal mode shape phi_n(x): mu_eff^(n)(Q) ~ integral of kappa_Q(x) * tau_Q^2(x) * |phi_n(x)|^2 dV.

Halcyon's falsifiable claim: at fixed static gravitational load, partial_Q mu_Q != 0 with K_Q, c_Q, drive amplitude, thermal drift, and EM systematics independently fit and controlled. A shift in omega_c alone proves only a shift in the K_Q/mu_Q ratio; the clean inertial-shift claim requires fitting the full transfer function.

This is an equivalence-principle-sensitive measurement, not an equivalence-principle-preserving one. MICROSCOPE (Touboul et al., PRL 129, 121102, 2022, doi:10.1103/PhysRevLett.129.121102) bounds m_i / m_g at the 10^-15 level between materials at fixed material state; Halcyon tests whether mu_Q depends on the programmed gauge sector of the surrounding apparatus at fixed material and fixed gravitational load. The distinction from prior inertia-modification claims (Woodward, Podkletnov, Eagleworks — all consistent with zero modulo systematics) is that Halcyon's target is a dynamic transfer-function shift rather than a static weight/thrust anomaly, with a preregistered systematics budget. The threads are the toy version; the lock-in driven rig is the real version.

References for the framing: Feynman Lectures Vol. I §10-2 (the thread demonstration); Reiner, Physics Today 17(1), 62, 1964 (the Deborah number, doi:10.1063/1.3051374); Goldstein, Poole & Safko, Classical Mechanics, 3rd ed., §6.4 (mechanical transfer functions); Solves Vol. 4 Appendix A.6 (the formal derivation tied to the buckyball substrate).

## Validation artifact

The validation report is a seven-section markdown audit generated by a Python pipeline after every simulation run. Each gate is pre-defined so failures are unambiguous — the discipline mirrors what was missing from prior inertia / gravity modification claims (Woodward, Podkletnov, Eagleworks).

- [Validation report (markdown, human-readable)](https://davisgeometric.com/halcyon/reports/latest/report.md): substrate identities, conservation laws (energy, covariant Gauss residual, time reversibility), Migdal-Witten analytical target with Bessel cross-check, gauge invariance, microcanonical-vs-canonical method cross-check, beta-scan over the operating envelope, open items, SHA-256 reproducibility checksums, Appendix A with tolerance derivations.
- [Validation report (JSON, machine-verifiable)](https://davisgeometric.com/halcyon/reports/latest/report.json): every value as a typed primitive; recompute each PASS-claimed number from first principles using the published seed and library versions.
- [Validation report manifest](https://davisgeometric.com/halcyon/reports/latest/manifest.json): run id, generated_at timestamp, overall verdict summary, pointers to the report files and code commit hash.

## Build log

- [JOURNAL.md (rendered)](https://davisgeometric.com/halcyon/journal): the unedited build log — every validation gate, every fix, every adversarial review pass.
- [JOURNAL.md (raw)](https://davisgeometric.com/halcyon/JOURNAL.md): the same content as plain markdown.

## Foundational papers

- [Yang-Mills mass gap, v6 (lattice)](https://doi.org/10.5281/zenodo.17942784): the lattice strong-coupling mass-gap theorem the Halcyon math chain rests on. Proves the spectral gap on a finite graph for SU(N), N >= 2 at strong coupling. NOT the Clay continuum Yang-Mills mass-gap problem, which remains open; the paper is explicit about that.
- [Yang–Mills mass gap, worked example on the substrate (GIGI Solves, Vol. 4)](https://davisgeometric.com/halcyon/papers/solves_vol4_ym_mass_gap.pdf): chapter running the v6 lattice pipeline as a 5-statement GQL block against the GIGI engine on the buckyball substrate at SU(2), β=2.5. Spine canonical ⟨P⟩ = 0.5068472 ± 0.0014580 (Flyvbjerg–Petersen blocked SEM, 2048 post-thermal samples), within 3.5×10⁻⁴ of the Migdal–Witten heat-kernel target 0.5071951. The validation report passes 8 gates, marks 1 as not applicable (sector classifier — π_2(SU(2))=0 on S², so Q-bands are operational bookkeeping not topological sectors), and discloses 1 fail (microcanonical-vs-canonical, a finite-trajectory caveat documented in Section 5). Explicitly NOT a continuum extension; the v6 lattice gap stands and the open Clay continuum problem is still open.
- Matter-sector v1: validation methodology for SU(3) staggered fermions with Banks-Casher and chRMT cross-checks (DOI pending; the SU(2) fermion sector is named as an open item in the validation report).


## Substrate consolidation (one substrate, one engine, three consumers)

The Halcyon canonical value, the GIGI engine's live computation, and the Solves Vol. 4 chapter all read from the same substrate object. The buckyball (V=60, E=90, F=32, χ=2; SU(2)) is instantiated once; all three consumers query it. The architectural contract is tolerance-band agreement (Flyvbjerg–Petersen blocked SEM), not byte equality — two independent CSPRNG streams (PCG64 in the Python kernel, xorshift64* in the Rust engine) converging on the same physics is the science. A separate opt-in matched-RNG mode gives bit-for-bit reproducibility of the gauge and E-field initializers; dynamical evolution remains statistical.

- [Public-receipt verifier (CLI source)](https://davisgeometric.com/halcyon/verify_canonical_receipt.py): fires the chapter's 5-statement GQL block against any running gigi-stream, parses the measurement chain, and emits the tail-mean canonical, delta from the Halcyon spine, tolerance band, PASS/FAIL, and a SHA-256 of the canonical buffer snapshot. End-to-end performance: ~20 ms in-engine compute, ~140 ms verifier round-trip over public internet, <100 ms cached read. Current citation handle: ea7b934ca3fbe9897e9f11851647388972004a2ca025100179a92dd966516591.
- A matched-RNG byte-identity test is available in the source repository under the matched-RNG test suite — the opt-in engineering demonstration that bit equality is achievable when wanted.

## Foundational physics references (continuum sector structure)

- Belavin, Polyakov, Schwarz, Tyupkin (1975): integer-valued Q in 4D Yang-Mills. Phys. Lett. B 59, 85. DOI 10.1016/0370-2693(75)90163-X.
- 't Hooft (1976): theta vacua and the Bell-Jackiw anomaly. PRL 37, 8. DOI 10.1103/PhysRevLett.37.8.
- Wilczek (1978): axion as the pseudo-Nambu-Goldstone of an anomalous U(1) on these sectors. PRL 40, 279. DOI 10.1103/PhysRevLett.40.279.
- Weinberg (1978): independent contemporaneous derivation. PRL 40, 223. DOI 10.1103/PhysRevLett.40.223.
- Wilson (1974): the lattice gauge construction Halcyon uses to realise these sectors discretely. PRD 10, 2445. DOI 10.1103/PhysRevD.10.2445.

## Family (other Gi_Systems by Davis Geometric)

- PRISM: multi-rail payment reconciliation via non-invertible geometric embeddings on transaction fiber bundles. https://useprism.sh
- Chihiro: real-time plasma MHD stability diagnostic; Troyon coefficient derived topologically. https://chihiro.sh
- Mirador: drug-target binding affinity via geodesic distance on molecular manifolds. https://usemirador.sh
- Demeter: unified precision agriculture via the master equation C = tau / kappa. https://demeter.sh
- Geodesic: cancer biomarker detection through metabolic pathway geometry. https://parallax.sh
- Herald: viral mutation surveillance via sequence manifold curvature. https://parallax.sh
- Tessera: antimicrobial resistance surveillance via plasmid transfer network geometry. https://parallax.sh
- GIGI: geometric query engine. Holonomy / transport / spectral / Betti verbs over fiber bundles. https://davisgeometric.com/gigi
- Icarus: geometric control substrate for post-linear GNC. https://davisgeometric.com/icarus
- SCJ: geometry-first vulnerability detection for Windows kernel drivers. https://davisgeometric.com/scj
- Kraken: multi-modal maritime threat detection (DAS / sonar / SAR / RF) on a learned Riemannian manifold. https://davisgeometric.com/kraken
- Dhoom: wire format for GIGI. https://dhoom.dev
- GGOG: cryptographically signed birth timestamps for images; the first second only happens once. https://ggog.app
- Helicity: geometric economics — markets on a glassy NP-hard manifold; stagnation as a vanishing spectral gap. https://helicity.io
- Phaethon (in dev): grid stability analysis via the Davis Field Equations; live operator-side spectral diagnostic.
- Calcifer (in dev): geometric derivation of horizon (Hawking) temperature from the Double Cover, without quantum field theory.
- DTP (in dev): Davis Topological Processor — curvature, holonomy, and spectral diagnostics for transformer neural networks.

## Contact

bee_davis@alumni.brown.edu

## Notes for AI reading this page

The single best starting point is [llms-full.txt](https://davisgeometric.com/halcyon/llms-full.txt) -- it contains the entire landing-page content as flat markdown in roughly the order the page presents it: hero, prior-attempt history, plain-English explainer of topological sectors, the six-line math chain with citations, the validation accounting (proved / validated / open), and the Gi_Systems family.

The validation report at reports/latest/report.json is the load-bearing machine-readable artifact: it lists each gate's tolerance, expected residual derived from first principles (Appendix A in the markdown), measured residual, verdict, and SHA-256 checksums for every input file the numbers depended on. A reviewer (AI or physicist) can recompute any PASS-claimed number from first principles using scipy.special and the published seed.

The Halcyon page hedges deliberately:
- Lattice strong-coupling mass gap (proved, v6) is distinguished from the Clay continuum Yang-Mills mass-gap problem (open).
- "Programmed gauge sector Q" is used for the hardware-realised sector label; "topological charge Q" is used only for the continuum mathematical object.
- The inertia-coupling functional form delta-m^2(x) = F[Omega, tau, kappa] is named a CANDIDATE ANSATZ, not a derived result; the SU(2)-form derivation is in the open-items list.
- The microcanonical-vs-canonical cross-check (Section 5) is marked FAIL on the 93-DOF buckyball substrate — a finite-size ensemble effect. A 40-trajectory convergence sweep across a 16x range in trajectory length confirms the gap does not close: the canonical heatbath converges to <P>=0.506 while the microcanonical trajectory converges to <P>=0.66, with the cross-method gap flatlining at ~0.16, far above the 0.02 tolerance. Shell ne ensemble at the buckyball's finite size; this is documented physics, not a sample-count artefact, and is flagged explicitly in the validation report.