What question did this study set out to answer?

The aim is to introduce a novel framework that unifies various physics-informed AI models to enhance dissipative intelligence across domains.

May 12, 2026Open Access

NEUROPIA: Neural Cognitive Field Unification for Cross-Domain Dissipative Intelligence

Key Points

The aim is to introduce a novel framework that unifies various physics-informed AI models to enhance dissipative intelligence across domains.
Developed the Neural Unified Propagator (NUP) as a 32-component operator integrating multiple physical systems.
Utilized Omni-Spectral Fourier Operator (O-SFO) and Grand Constraint Network (GCN) to enforce conservation laws in the architecture.
Conducted validation using eight canonical multi-physics benchmarks.
Achieved a 96.8% mean Unified Efficiency Index across benchmarks.
Observed a 91.4% mean reduction in cross-domain dissipation rates.
Demonstrated a 12.3× instability suppression factor compared to uncontrolled baselines.

Abstract

NEUROPIA (E-LAB-10) is the tenth and culminating Physics-Informed Artificial Intelligence (PIAI) framework of the EntropyLab research program. The framework introduces the Neural Unified Propagator (NUP), a 32-component cross-domain operator that unifies the dissipation channels of all nine predecessor EntropyLab projects — magnetohydrodynamics (MAGNA-FLOW), thermodynamics (THERMO-NET), quantum optics (PHOTON-Q), curved spacetime (GRAVI-NEURAL), thermodynamic engines (ENTRO-ENGINE), reactive chemistry (CHEM-ENTROPIA), biological metabolic networks (BIO-ENTROPIA), and AI inference thermodynamics (ENTRO-AI) — into a single learnable architecture governed by the Generalized Dissipation Action. Three mathematically rigorous constructs power the framework: (1) the Omni-Spectral Fourier Operator (O-SFO), a 32×32 complex cross-domain spectral kernel operating in unified frequency space; (2) the Grand Constraint Network (GCN), which enforces all nine EntropyLab conservation laws — helicity, entropy monotonicity, Onsager reciprocity, Bianchi identity, and quantum unitarity — as hard architectural priors; and (3) the Unified Flux Resolver (UFR), a model-predictive control engine tracking the Generalized Stress-Energy Tensor across all coupled subsystems. Validation across eight canonical multi-physics benchmarks demonstrates a 96.8% mean Unified Efficiency Index, a 91.4% mean cross-domain dissipation reduction, and a 12.3× instability suppression factor relative to uncontrolled baselines, approaching the theoretical multi-physics entropy floor within 3.2%. DOI: 10.5281/zenodo.20092199 | MIT License | Entropy (MDPI), ISSN 1099-4300 | EntropyLab Program E-LAB-10 | ORCID: 0009-0003-8903-0029

Read Full Paperexternally

KI fragen

Bookmark

View Full Paper