Over the past several years I have been building a substantial formal research program — machine-verified, mathematically precise, and published with permanent DOIs on Zenodo. Today I am making the full index available at novaspivack.com/research. This post is an introduction to what the program covers and why I think it matters.

Update, May 2026: Since this post was first published in April 2026, the program has grown significantly. The UGP Physics program — the quantitative physics layer of Reflexive Reality — is now published as 28 papers (P00–P27), each with a permanent Zenodo DOI. A new section below covers the UGP Physics program in full.

The Central Question

What must any consistent, self-contained reality look like? If a universe has no “outside” — no external model-selector, no transcendent lawgiver, no escape from the closure of its own laws — what can we prove about its structure, purely from that fact?

This is not a philosophical question in the hand-waving sense. It turns out to be a mathematical question — one that can be approached with the same rigor as a theorem in logic, and machine-checked in a proof assistant. That is exactly what the NEMS program does.

The NEMS Program: No External Model Selection

NEMS — No External Model Selection — is the core of the research program. It consists of 93 papers (plus companions), all machine-verified in Lean 4 with a strict zero-sorry, zero-custom-axiom policy. The central principle is simple: a universe with no outside cannot have its laws, parameters, or ontological ground selected by anything external. Everything must come from within.

What follows from this single commitment is remarkable. Among the results the program proves:

• The structure of gauge theory is forced. The axioms of Perfect Self-Containment (PSC) narrow the space of admissible 4D quantum field theories to SU(3)×SU(2)×U(1) with anomaly-minimal chiral matter and exactly three generations — not as an empirical coincidence, but as a theorem.
• The Born rule is the unique consistent probability assignment. In any perfectly self-contained theory with records carrying quantum effect structure, there is exactly one normalized, POVM-additive probability assignment compatible with closure: the Born rule. The uniqueness is fully machine-checked — no other assignment is consistent with PSC. Paper 14 closes the reverse direction: Born-internal complete semantics implies PSC, making the equivalence bidirectional.
• Determinism is forbidden. Diagonal-capable physical records forbid total-effective record determinism. A closed universe cannot be fully deterministic — not because of quantum randomness, but because of a deeper structural constraint.
• Non-algorithmic adjudication is necessary. Standard computation cannot resolve the indeterminacies forced by closure. Something beyond computation — what the program calls Transputation — is structurally required.
• The Simulation Hypothesis and Block Universe are refuted. The universe must be actively run from within — no total-effective static algorithm can perfectly emulate its internal adjudication. If such an emulation existed, it would produce a computable decider for record-truth, directly violating the diagonal barrier. Consequently, internal adjudicators — agents — are not biological accidents but the necessary execution engine of physical reality. The universe is not a simulation; it is a process that cannot be replaced by any description of itself.
• Physical incompleteness from universal computation. Any perfectly self-contained physical theory that contains universal computation is physically incomplete — a machine-checked diagonal barrier showing that closure and computational universality together force undecidable physical facts.
• The arrow of time follows from closure. Stable physical records and the no-overwrite constraint force the irreversibility of time as a theorem, not an assumption.
• Syntax cannot exhaust semantics. In any reflexive system, the syntactic layer cannot fully capture its own semantic content. There is always an irreducible semantic remainder — a formal no-reduction theorem for reflexive systems that closes the gap between Gödel-style incompleteness and the deeper question of meaning.
• Closure without exhaustion. Reflexive systems close without collapsing: the closure of a self-referential system does not exhaust its semantic structure. Internal semantic non-exhaustion is a theorem, not a conjecture — and it means that no reflexive system can fully describe itself, even in principle.
• Representational incompleteness. No parametric self-model can cover its own diagonal. For any attempted complete internal self-representation, there is always a blind spot — a result that closes six distinct “no-escape” routes (regress, collapse, cross-object readout, partial models, decidable self-reference, model substitution) as theorems.
• The ontological ground of reality is necessary and non-null. Paper 63 proves that if nontrivial reflexive reality exists, there must be a necessary pre-categorial ontological ground (Alpha) — not contingent, not derivable from syntax, not nihilistically empty. Paper 55 proves that qualia are irreducible semantic ledger content: the “hard problem” of consciousness, as traditionally posed, is a category error — it demands syntax alone generate qualia from outside the ledger, which is structurally impossible. Paper 70 (the Golden Bridge) unifies ground, articulation, and manifestation-in-awareness as coordinated irreducible aspects of one primordial ontological fact. The program does not merely argue these claims — it machine-checks them.

The program also covers quantum gravity, black hole information, holography, the arrow of time, institutional epistemics, self-improvement under diagonal constraints, and much more — 93 papers in all, organized into thematic groups. Every result is machine-checked. Every claim is a theorem.

The Universal Generative Principle (UGP) Physics Program

The Universal Generative Principle (UGP) is a deterministic, arithmetic framework in which the specific numerical values of the Standard Model’s physical parameters — masses, coupling strengths, quantum numbers — are not free inputs but forced by the rules of integer arithmetic applied to a closed, self-referential system. Where NEMS establishes why the Standard Model structure is the only consistent option for a self-contained universe, UGP asks and answers the harder question: what exact numbers must that structure have?

The mechanism works as follows. Three axioms — locality, symmetry, and minimum description length — uniquely select a single integer seed: the canonical triple (1, 73, 823). This is not a parameter choice; it is a forced consequence of arithmetic, machine-certified as the unique lexicographically minimal survivor of a two-stage sieve. From this seed, a cascade called the Generative Triple Evolution (GTE) map unfolds a rigid three-step orbit — (1, 73, 823) → (9, 42, 1023) → (5, 275, 65535) — propagating exact bare gauge-coupling rationals, canonical fermion triples, and quantum-number assignments with no free parameters.

The deeper insight is geometric: UGP identifies a lower-dimensional constraint manifold — an arithmetic spine — that the Standard Model’s 25-dimensional parameter space must lie on. The observed parameters are not arbitrary points in that high-dimensional space; they sit in a rigid algebraic groove forced by self-consistency. For the non-physicist: imagine 25 independent sliders, each settable to any value, giving an enormous space of possible physics. UGP finds the single groove they all must fall into, and shows the observed values are already there.

The program spans 28 papers (P00–P27) covering particle physics, nuclear structure, molecular biology, ecology, information theory, and cosmology — all derived from the same arithmetic seed. The core Category-A derivation chain is machine-checked in Lean 4 across ugp-lean and ugp-physics-lean (118+ modules, zero sorry, standard Mathlib axiom signature). Key results:

Standard Model parameters derived from first principles

• Standard Model parameter spectrum derived from first principles (P01). Three axioms — locality, symmetry, and minimum description length — uniquely select a single integer seed, and a rigid arithmetic cascade from that seed generates the Category-A structural backbone of the Standard Model parameter spectrum: exact bare gauge-coupling rationals, canonical fermion triples, and quantum-number assignments with no free parameters. The framework is complementary to quantum field theory — it fixes the numerical constants that a QFT Lagrangian currently takes as experimental inputs, leaving QFT to supply the scattering dynamics. The core proof path is machine-certified in Lean 4 with zero sorry and zero custom axioms.
• Standard Model uniqueness confirmed by exhaustive universe enumeration (P12). An exhaustive enumeration over 34,560 candidate universes across 12 gauge groups — including Pati-Salam and E₆ — places the Standard Model at rank #1, with all beyond-Standard-Model alternatives decisively eliminated; fewer than 0.03% of survivors share the SM signature. This is confirmed independently by three rigidity mechanisms (PSC theory-space selection, UGP arithmetic, and seed rigidity) that jointly select the same canonical physical branch under an explicit and audited premise bundle.
• Standard Model gauge group uniquely derived (P06). The algebraic and geometric structure of the UGP framework is shown to force SU(3)×SU(2)×U(1) as the unique consistent gauge group. The derivation proceeds through explicit construction of the color and electroweak fiber structure from GTE orbit data, recovering the hypercharge table, gluon confinement, and electroweak symmetry breaking from first principles.
• SM gauge group uniqueness certified by exhaustive computation (P14). A complementary computational certificate confirms PSC-selected Standard Model structure through three new parameter-free coupling-ratio predictions and a full constraint independence analysis. The formal and computational results independently converge on the same canonical physical branch.
• Standard Model uniqueness proved by three independent rigidity mechanisms (P05, P12). Three independently developed rigidity results — from PSC theory-space selection, UGP arithmetic constraints, and topological stability — jointly select the same canonical physical branch from the enormous prior space of possible theories. Fewer than 0.03% of surveyed alternative universes share the SM signature, and the rigidity mechanisms are shown to be logically independent proofs of the same uniqueness claim.

Particle spectrum and quantum numbers

• Complete particle spectrum reconstructed (P02). A large-scale discovery run generates 1,000,035 candidate particles below 173 GeV from the GTE orbit; all 24 known SM particles rank in the top viability tier with 50-fold enrichment compared to a random draw (p < 10⁻⁴). Nine genuinely novel GTE candidate states (GTE-P1 through P11) emerge as falsifiable mass-band predictions for future experiments, with calibration uncertainties of roughly 1–40%.
• SM topology derived from braid structure (P17). The quantum numbers of the Standard Model — spin, charge, color, family — are derived from an information-theoretic braid structure called the Braid Atlas. The winding, chirality, hypercharge, and color-fiber data of the SM are shown to be uniquely determined by the topological constraints of the GTE orbit, recovering the full fermion and gauge-boson quantum-number tables from braid data.
• Interaction skeleton theorem machine-verified (P22). In Lean 4, with zero sorry for all theorem-grade claims, the UGP winding, chirality, hypercharge, and color-fiber data are proved to determine the Standard Model’s finite renormalizable vertex structure exactly — the Silver closure (UGP vertices ↔ SM vertices) and Gold closure (UGP Yukawa schemas = SM Yukawa schemas) are both machine-verified.
• Neutrino mass-squared ratio predicted (P21). The neutrino mass-squared ratio Δm²₁₂/Δm²₂₃ is derived from the Braid Atlas structure — the first parameter-free structural prediction of this quantity in the literature. The prediction is consistent with current experimental bounds at 0.16σ.

Mass relations and the cyclotomic substrate

• Koide relation elevated to structural theorem (P18). The Koide identity — a 1983 phenomenological formula connecting the three charged-lepton masses that has had no accepted theoretical origin — is promoted to a theorem: the Koide phase θ = 2/9 follows unconditionally from QCD’s color rank N_c = 3 via a strand-count identity, machine-checked in Lean 4. The resulting closed form predicts the tau-lepton mass to 61 parts per million.
• Charged-fermion mass spectrum unified in a cyclotomic-12 structure (P19). Two structural mass relations reduce the nine charged-fermion masses to two empirical scale inputs, with the underlying geometry living in the cyclotomic-12 field Q(ζ₁₂). This identifies a common algebraic home for all charged-fermion masses, connecting the Koide result to the broader GTE arithmetic structure.
• Nuclear magic numbers analytically derived (P03). All seven nuclear magic numbers {2, 8, 20, 28, 50, 82, 126} are derived analytically from the GTE cascade via one-pion-exchange physics, with the pion parameters themselves predicted rather than fitted. The spin-orbit coupling produces all seven as large energy gaps in the single-particle spectrum — a complete order-of-magnitude derivation of nuclear shell closure from pure arithmetic.

Cross-domain selection principles

• Information Profit Threshold derived and empirically validated (P15). Any self-referential system must generate information faster than it dissipates it; making this precise yields the Information Profit Threshold (IPT ≈ 1.1309), derived from the Reflexive Landauer Bound with the exact value following from the golden ratio φ. The threshold is validated across eight independent real-world tests: the published photosynthesis-to-respiration ratio for tropical moist forest ecosystems is 1.130 — matching IPT to four decimal places — and the boundary separates thriving from declining ecosystem types with no free parameters.
• Standard genetic code uniquely selected (P25). A two-stage sieve — wobble admissibility (reducing the search space by a factor of 10⁵⁰) followed by eight biological viability criteria — is applied to all possible codon-to-amino-acid mappings. Out of 10⁵ randomly generated codes, none outperforms the standard genetic code under all eight criteria jointly, and it ranks at z = +3.22σ (p < 0.009%). The standard genetic code is not merely good — it is the unique survivor.
• General Theory of Selection: one mechanism across all domains (P26). The same two-stage sieve that selects the Standard Model — admissibility followed by viability against the IPT — operates identically across nuclear physics, prebiotic chemistry, ecology, corporate survival, and microbial metabolism. The IPT value recurs empirically across all five domains (microbial metabolic flux correlation r = 0.971, p < 0.001), and the same algebraic substrate Q recurs independently in the SM parameter spectrum, E8 integrable QFT mass ratios, and WZW quantum dimensions. A single mathematical mechanism underlies the emergence of stable structure across complex systems at every scale.

Dynamics, meta-laws, and the computable universe

• Nine universal meta-laws unified (P07). The UGP/GTE framework is shown to subsume the premises of nine universal meta-laws of physics (ML-1 through ML-9) — principles such as the universality of thermodynamic limits, holographic bounds, and the emergence of classical behavior. The UGP provides a common vocabulary and derivational basis for these meta-laws, showing they are consequences of the same arithmetic structure rather than independent postulates.
• Ontological Dissonance Minimization identified as generative meta-law (P11). Minimizing a dissonance functional D — the degree to which a system’s self-model conflicts with its actual dynamics — is shown to produce results empirically consistent with maximizing a spatial-integration measure across the PR-0 simulation. This identifies a new candidate for the deepest generative principle underlying physical law.

Cosmology and the Self-Referential Renormalization Group

• Black hole information paradox addressed within the framework (P16). Hawking evaporation, modeled as a Lindblad master equation, is shown to be unitarily equivalent to reversible evolution via the Stinespring dilation theorem. The MFRR framework provides a constructive mechanism (the PT⁻¹ Transputation operator) identifying which specific unitary implements the information recovery; numerical verification achieves Stinespring fidelity F > 1 − 10⁻⁸, consistent with early-time Page curve dynamics.
• Self-Referential Renormalization Group identifies unique physical fixed point (P27). The SRRG constructs a gradient-flow theory on the space of self-referential physical theories, proving that any such theory flows to a unique IR fixed point S* with minimal U(1) symmetry and contraction rate 1/φ. At this fixed point, the IPT (1.1309) emerges as the efficiency ratio. The β-function β(λ) = (λ − φ)(λ − 2), machine-certified in Lean 4, yields conditional derivations of θ_QCD = 0 (addressing the strong CP problem), three fermion generations, and Higgs quartic coupling recovery.

The complete program — 36 citable records including papers, Lean archives, and corpus bundle — is available at the UGP Physics hub on Zenodo.

Extended Mathematical Programs

Surrounding the NEMS core is a family of extended programs, each proving deep structural results about self-referential and reflexive systems:

• Infinity Compression — Proves that canonical certification cannot exhaust reflective structure. When a formal system has a bare certification layer and a richer realization layer, the two cannot be collapsed. The difference is organized by fibers, sections, and obstruction laws. The summit theorem — Reflective Non-Exhaustion — is a unified fixed-point result connecting the internal collapse barrier to the external realization gap: the enriched structure above the bare carrier is not just different, it is irreducibly richer in a precise, machine-checked sense. Seven papers, all machine-checked.
• Reflexive Architecture — A machine-checked synthesis of NEMS, Abstract Indexed Programming Systems, and Infinity Compression into a unified layered architecture. Two summits: one on closure and realization, one on what maps forget.
• Representational Incompleteness — No parametric self-model can cover its own diagonal. Six adversarial “no-escape” routes are closed as theorems.
• Reflective Fold Obstruction — If a predicate is preserved along primitive steps of an internal relation, the internal reflexive-transitive closure cannot reach any state falsifying that predicate. Fold barriers are real and structurally precise.
• Observer Non-Exhaustibility — Classifies all internal strategies for “exhausting” observer architectures into three provably blocked families, with a positive non-collapsing residual from the awareness arc of the reflexive-closure program.
• Adequacy Architecture and Reflexive Architecture Nonexhaustibility — Outer admissibility, certificate worlds, the general science of reflexive systems, and the Reflexive Development Law.

Novelty Theory

A separate program — independent of Reflexive Reality — investigates a different kind of structural constraint. Novelty Theory proves that fixed deterministic laws do not entail final explanatory closure. Even under perfectly lawful generators, there exist phase towers that no fixed admissible reducer can close. At the crown, later regimes become required for structural truths about the generator itself. This is not Gödelian incompleteness, not computational irreducibility — it is explanatory anti-closure under exact generation. A machine-checked result about why genuine novelty is not just possible but structurally necessary.

Foundational Monographs

Four book-length monographs underpin the program. The Mathematical Foundations of Self-Referential Systems develops a unified mathematical treatment from computability through transfinite and field-theoretic structure — Recursive Representation Theory, the Self-Referential Renormalization Group, information geometry, and the Self-Computation Principle. The Self-Defining Universe develops the formal theory of perfect self-containment and the meta-topological conditions under which mathematical and physical frameworks can co-emerge from closure. The UGP Foundational Monograph (P08) extends the foundational framework into the physics domain, developing the GTE map and its derivation of Standard Model structure from first arithmetic principles. The Mathematical Foundations of Reflexive Reality (MFRR, P13) develops the theoretical bridge between the Reflexive Reality program and the UGP Physics derivations — accompanied by three companion papers providing a reader’s guide, a gentle popular-science introduction, and a technical summary for physicists. All monographs are available on Zenodo.

Why Machine Verification?

Every major result in this program is machine-checked in Lean 4, an interactive theorem prover built on dependent type theory. This means the proofs are not informal arguments — they are formal derivations that a computer has verified are logically correct, given the stated hypotheses and without any sorry placeholders or custom axioms. The Lean source code for every paper is publicly available on GitHub.

Why does this matter? Because the claims are bold. The Standard Model from first principles. The necessity of non-algorithmic processing. A formal theory of consciousness. Claims this strong invite skepticism — and they should. Machine verification is the response to that skepticism: every inference step is checked, every dependency is explicit, every assumption is declared. The work can be audited by anyone with a Lean installation.

What Is Published

All of the following are publicly available with permanent Zenodo DOIs:

• 93 NEMS physics papers (standalone PDFs, one DOI each)
• 28 UGP Physics papers (P00–P27), each with a Zenodo DOI
• 7 Infinity Compression papers
• 20 Lean 4 software archives (17 NEMS + 3 UGP Physics: ugp-physics-lean, srrg-lean, lean-ugp)
• 1 NEMS corpus bundle (all 93 NEMS papers as a single archive)
• 1 UGP Physics corpus bundle (all 28 UGP papers and archives)
• 1 Reflexive Reality corpus bundle (all 142 papers from NEMS and UGP Physics)
• 1 PSC computational validation dataset
• 4 foundational monographs (2 NEMS + P08 UGP Foundational Monograph + P13 MFRR)
• 3 MFRR companion papers (reader’s guide, gentle introduction, technical summary)
• The Novelty Theory paper and Lean archive
• 1 NEMS program hub on Zenodo with hasPart / isPartOf DataCite metadata (123 NEMS records)
• 1 UGP Physics hub on Zenodo with hasPart links to all 36 UGP Physics records
• ~159 citable records total across both programs

The complete index, with abstracts, DOI links, and GitHub links for every item, is at novaspivack.com/research. All Lean source repositories are public at github.com/novaspivack.

Where to Start

If you are new to the program, I recommend starting with one of these entry points depending on your background:

• General reader: Overview of the NEMS Framework (Paper 00) — the suite overview, designed to be accessible.
• Physicists: The NEMS Framework for Physicists (Paper 07) — covers gauge structure and the Standard Model derivation.
• AI / AGI researchers: Reflexive Reality: A Survey for AI and AGI Researchers (Paper 89) — theorem-grade constraints on intelligence, agency, and machine consciousness.
• Mathematicians / logicians: A Survey for Formal Theory Specialists (Paper 87) — new impossibility results and proof methods.
• Big picture: Big Results from the NEMS Program (Paper 81) — flagship theorems across all domains.
• Flagship result: Closure Without Exhaustion (Paper 91) — a theorem of internal semantic non-exhaustion for reflexive systems.
• Physicists / physics-curious: Survey and Reader’s Guide to UGP Physics (P00) — covers all 27 UGP derivations with annotated reading paths for physicists, mathematicians, and philosophers.

The full research index is at novaspivack.com/research. Comments, questions, and critical engagement are welcome.