Work

Consolidated P–NP: Source-Native Exact Boolean Computation, Adaptive Hypergraph Dynamics, and Information Thermodynamics — Mathematical Foundations, Computational Verification, and Lean Certification

Year: 2026
Type: preprint
Abstract: This research publication consolidates a multistage mathematical, computational, and theoretical-physics investigation of the P versus NP problem, exact Boolean satisfiability, model counting, and the construction of computationally economical representations from arbitrary source constraints. The central objective is to determine whether the known polynomial-time verification of Boolean witnesses can be complemented by a universally efficient, deterministic, source-native decision constructor. The notation $NP'\rightarrow P'$ identifies this construction objective within the research program, rather than a separately defined standard complexity class. For a Boolean formula $F$ over $n$ variables and a terminal or continuation constraint $C$, the exact source response is \[Z(F,C)=\sum_{x\in\{0,1\}^{n}}\mathbf{1}[F(x)]\mathbf{1}[C(x)],\qquad\operatorname{SAT}(F\land C)\Longleftrightarrow Z(F,C)>0.\] The investigation develops a unified mathematical treatment of signed forward–reverse computation, Boolean source elimination, Walsh–Hadamard transformations, circle inversion, Bessel–Cayley correspondence, affine and nonlinear source grammars, finite-state transfer recurrences, graph and hypergraph topology, parity and holonomy invariants, adaptive boundary conditions, and dynamic-to-static computational mappings. A central exact identity connects forward and reverse boundary responses through signed spectral contraction: \[\sum_{S\subseteq[w]}\widehat{L}(S)\widehat{R}(S)=2^w\sum_{x\in\{0,1\}^{w}}L(x)R(x).\] These identities permit exact source-preserving transformations and certified calculations within their stated mathematical domains. The research further examines the evolving topology of source dependencies, connected residuals, separator interfaces, independent cycles, reusable certificates, and contextual equivalence. Particular emphasis is placed on **full future-context coverage**, rather than reducing computational complexity to initial seed number, apparent representation size, or isolated operation counts. The accompanying mathematics–physics analysis develops an information-theoretic and statistical-mechanical correspondence between Boolean constraints, weighted source energies, partition functions, Shannon entropy, negentropy, and free-energy functionals. For a nonnegative source energy $E_F(x)$ whose zero-energy configurations are precisely the satisfying assignments, \[\mathcal{Z}_{\beta}=\sum_{x\in\{0,1\}^{n}}e^{-\beta E_F(x)},\qquad\lim_{\beta\rightarrow\infty}\mathcal{Z}_{\beta}=\#\operatorname{SAT}(F).\] The entropy conversion between natural logarithmic units and binary information is \[S_{\mathrm{thermo}}=k_B H_{\mathrm{nat}}=k_B\ln(2)\,H_{\mathrm{bits}}.\] These relationships provide an exact thermodynamic interpretation for specified source-energy models and motivate information-based analysis of computational work, reuse, and representation compression. They do not, independently, establish polynomial-time algorithms. Computational economy is examined through complete chronological accounting of source discovery, construction, evaluation, verification, reuse, and decoding. A modular exact-counting result shows that a supplied division-free integer circuit computing $\#\operatorname{SAT}(F)$ may be evaluated modulo $2^{n+1}$ without losing its bounded final count, thereby distinguishing intermediate numerical magnitude from the unresolved cost of constructing a sufficiently economical circuit. The consolidated program includes exact general identities, polynomial algorithms for identified structured Boolean classes, conditional complexity bounds, and computational and formal-verification evidence. The audited public-safe reproducibility archive contains 53 files and reports successful execution of ten independent Python verification suites and fresh compilation of five Lean theorem modules. Recorded checks include 2,040 cyclic-source target instances, 36 nonlinear source-lift records, 20 scheduler decisions, 13 source-credit graph tests, six spectral interfaces, 288 CNF/restricted-context evaluations, and 112 signed-Walsh comparisons. The archive preserves the scope and limitations of these tests, including historical outcomes that cannot be fully independently reconstructed from the public materials. The publication also examines proposed subexponential contraction and Exponential Time Hypothesis (ETH) refutation approaches, distinguishing exact algebraic reductions, finite computational successes, conditional asymptotic conclusions, and unresolved general-case complexity guarantees. The master mathematics–physics exposition is supported by detailed derivations, computational prove-out documentation, theorem-to-evidence cross-references, Lean source records, Python validation programs, and reproducibility materials. Enhanced internal discovery-policy implementations are outside the scope of the public release. The resulting contribution is an integrated, auditable framework connecting exact Boolean computation, adaptive source networks, information-theoretic energetics, and formal mathematical certification. The unrestricted $NP'\rightarrow P'$ construction, a universal polynomial-time SAT algorithm, $P=NP$, and unconditional ETH refutation remain open proof obligations rather than established conclusions.
Language: English
Open Access status: green
OpenAlex ID: W7235257928
Cites:
Cited by:
Related to: 10