Abstract: Large Language Models (LLMs) have become foundational to modern Artificial Intelligence (AI), enabling advanced reasoning, multimodal understanding, and scalable human-AI interaction across diverse domains. This survey provides a comprehensive review of major proprietary and open-source LLM families, including GPT, LLaMA 2, Gemini, Claude, DeepSeek, Falcon, and Qwen. It systematically examines architectural advancements such as transformer refinements, mixture-of-experts paradigms, attention optimization, long-context modeling, and multimodal integration. The paper further analyzes alignment and safety mechanisms, encompassing instruction tuning, reinforcement learning from human feedback, and constitutional frameworks, and discusses their implications for controllability, reliability, and responsible deployment. Comparative analysis of training strategies, data curation practices, efficiency optimizations, and application settings highlights key trade-offs among scalability, performance, interpretability, and ethical considerations. Beyond synthesis, the survey introduces a structured taxonomy and a feature-driven comparative study of over 50 reconstructed LLM architectures, complemented by an interactive visualization interface and an open-source implementation to support transparency and reproducibility. Finally, it outlines open challenges and future research directions related to transparency, computational cost, data governance, and societal impact, offering a unified reference for researchers and practitioners developing large-scale AI systems.

Abstract: Climate change increases uncertainty in agricultural production and rural livelihoods, encouraging farms to pursue diversification strategies that can buffer climate-related risks. At the same time, the growing use of digital and AI-based climate and decision-support tools raises questions about how the transparency of such information shapes farm-level adaptation. This study examines the relationships among AI transparency, climate awareness, decision confidence, agritourism diversification intention, and perceived farm resilience in climate-sensitive rural systems. Data were collected through in-person fieldwork conducted throughout 2025 among agritourism-oriented farm operators in two Serbian rural clusters: a Western mountain agritourism belt and an Eastern/Southeastern dry-stress zone. Using structural equation modeling, the analysis reveals a coherent pattern of positive associations across all modeled relationships. Higher perceived transparency of AI-based climate information is associated with stronger climate awareness, greater decision confidence, increased intention to diversify toward agritourism, and higher perceived farm resilience. Perceived farm resilience was most strongly related to agritourism diversification intention, underscoring diversification as a key adaptive pathway under climate stress. The findings highlight AI transparency as a critical informational precondition for adaptive decision-making and resilience building, with implications for farmer-centric digital tools and rural climate adaptation policy.

Abstract: Background/Objectives: Obesity and type 2 diabetes are increasingly common among patients undergoing hip and knee arthroplasty and are associated with higher risks of prosthetic joint infection, impaired wound healing, and prolonged hospitalization. Dietary carbohydrate restriction has demonstrated benefits in glycemic control and weight reduction, but its feasibility and safety in the perioperative arthroplasty population remain underexplored. This pilot study evaluated the safety, feasibility, and short-term metabolic effects of a low-carbohydrate diet supported by WhatsApp-based meal photo-logging in patients undergoing total hip or knee arthroplasty. Methods: A retrospective cohort analysis was performed on 43 patients enrolled in a carbohydrate-restricted dietary programme between 2021 and 2024. Patients submitted photographs of all meals via WhatsApp with a minimum contact frequency of four times daily, enabling real-time feedback and medication adjustment. Anthropometric and metabolic parameters, including weight, BMI, HbA1c, renal function, and lipid profile, were assessed before and after the intervention. Results: Participants (mean age 69.12 ± 7.51 years) demonstrated significant improvement across several metabolic markers. Mean weight decreased by 5.74 kg (p < 0.001), BMI by 2.26 kg/m² (p < 0.001), and HbA1c by 0.72% (p < 0.001). No episodes of severe hypoglycemia or perioperative discharge delays related to glycemic instability were observed. Renal function remained stable, with no significant change in eGFR (p = 0.442). Among patients with available lipid data, LDL-cholesterol and total cholesterol increased, while triglycerides showed a non-significant downward trend. Conclusions: A low-carbohydrate diet combined with high-frequency digital monitoring appears feasible and safe in an elderly arthroplasty population, achieving meaningful short-term improvements in weight and glycemic control without adverse renal or hypoglycemic events. The lipid changes observed warrant cautious interpretation. Larger prospective studies are needed to confirm the clinical impact of this approach and its relevance to perioperative optimization.

Abstract: This paper proposes an innovative vision-language model (VLM) driven predictive platform that synergistically integrates swarm robotics coordination with post-quantum digital signatures to enable fully autonomous navigation for green vessels eco-friendly ships leveraging hybrid renewable propulsion systems such as biofuels, hydrogen fuel cells, and wind-assisted technologies. Traditional maritime navigation systems struggle with dynamic oceanic conditions, including unpredictable weather patterns, high-traffic congestion, and escalating cyber threats, which compromise supply chain efficiency and sustainability goals. The proposed framework addresses these challenges by deploying a multimodal VLM core that fuses real-time visual data from onboard LiDAR, infrared cameras, and radar with textual inputs from AIS (Automatic Identification System) broadcasts, satellite weather forecasts, and nautical charts. This fusion generates interpretable probabilistic predictions of future states, such as wave-induced trajectory deviations or collision risks, enabling proactive rerouting that minimizes hydrodynamic drag and emissions.Swarm robotics augments individual vessel autonomy through decentralized fleets of unmanned surface vehicles (USVs) that dynamically form protective convoys or scouting formations, optimizing collective energy use via bio-inspired particle swarm optimization conditioned on VLM outputs. To safeguard against quantum computing vulnerabilities inherent in classical RSA or ECC protocols, the platform embeds lightweight Dilithium or Falcon post-quantum signatures for authenticating sensor streams, cargo manifests, and inter-vessel commands, ensuring non-repudiation even under harvest-now-decrypt-later attacks. Validation occurs within high-fidelity maritime digital twins that simulate full-scale operations, incorporating computational fluid dynamics for vessel hydrodynamics and stochastic perturbations for resilience testing.Extensive simulations demonstrate transformative performance fuel savings exceed 38% via predictive eco-routing, collision avoidance precision reaches 97.2% in dense fog scenarios, and cryptographic overhead remains below 5% bandwidth utilization on edge-constrained USVs. Supply chain resilience improves markedly, with recovery time from simulated disruptions reduced by 52%, fortifying global logistics against climate volatility and geopolitical risks. This work pioneer’s end-to-end integration of VLMs, swarms, and quantum-safe primitives, laying a robust foundation for scalable, secure, and sustainable autonomous maritime ecosystems aligned with UN Sustainability Development Goals.

Abstract: Considered the possible assembly malfunction in control loop, this paper researches the sliding mode observer(SMO) design for a linearized physical system with environmental disturbances and sensor faults in some constrained conditions on system structure, to set up the fault detection and isolation(FDI) scheme for system in the loop. On one hand, by utilizing the features of fault distribution, the coefficients of fault and disturbance in unobserved subsystem are canceled by state transform under the presumed conditions. Then SMO served in FDI for observable subsystem is constructed, where the convergence of observed error is verified by analysis on Lyapunov functions. On the other hand, for the general situation when fault and disturbance are distributed randomly, the coefficients of fault corresponding to unobserved states are canceled by imposing some similarity transforms on system matrix, such that a reconfigured SMO is designed to counteract and detect the fault in observable subsystem. Furthermore, using inequality transform, the convergence of observed error is shown to be bounded with oscillation, which is proved for the existence of disturbance. Finally, FDI scheme is applied and tested in a fixed-wing airplane system to validate the stability of SMO.

Abstract: Nighttime vehicle detection poses significant challenges due to reduced visibility, uneven illumination, and increased noise in low-light imagery. While deep learning approaches have achieved remarkable success in daytime scenarios, their application to nighttime conditions remains constrained by the scarcity of specialized datasets and the computational demands of existing architectures. This paper presents three primary contributions to address these challenges. First, we introduce the Low-light Vehicle Annotation Dataset (L-VAD), comprising 13,648 annotated frames captured exclusively during nighttime conditions across three vehicle categories: motorcycle, car, and truck/bus. Second, we propose TinyNight-YOLO, an ultra-lightweight detection architecture achieving competitive performance with only ∼1.0 million parameters—representing a 2.6× reduction compared to YOLO11- N and 26.4× reduction compared to YOLO11-L. Third, we provide a comprehensive benchmark evaluating ten model variants across YOLO11 and YOLOv12 families. Experimental results demonstrate that TinyNight-YOLO achieves F1-Score of 0.9207 and mAP@50 of 0.9474, representing only 1.44% accuracy reduction compared to models 2.6× larger, while outperforming YOLOv12-L (26.4M parameters) despite having 26.4× fewer parameters. Among full-scale models, YOLO11-L achieves the highest F1-Score (0.9486), while YOLO11-M attains superior mAP@50-95 (0.7271). The L-VAD dataset is publicly available at Mendeley Data (doi: 10.17632/h6p2w53my5.1), providing the research community with a dedicated resource for advancing nighttime vehicle detection. The proposed TinyNight-YOLO architecture enables practical deployment on resource-constrained edge devices while maintaining detection accuracy above 94% mAP@50.

Abstract: The challenging goal of equipping HF radars with a target classification ability has been pursued for many years, yet no satisfactory system-level methodology has been reported. This shortcoming severely limits the utility of radar information as, without knowing the nature of detected objects, there is little prospect of understanding the situation and tailoring a suitable response. In this paper, we present a framework within which a comprehensive approach to target characterization can be formulated. We proceed to explore a wide range of physical mechanisms whereby target information is impressed on HF radar echoes, illustrated with real data. The paper concludes with a commentary on the difficulty of integrating target classification, recognition and identification procedures with other radar tasks and resource management.

Abstract: Although researchers document the political utility of cabinet reshuffles in African presidential systems extensively, they devote insufficient empirical attention to the trade-offs these reshuffles impose on economic performance. This paper delivers robust evidence that frequent ministerial changes inflict substantial, often underestimated costs on economic growth. The analysis draws on data from 19 African nations spanning 2006 to 2023 and applies a polynomial dynamic panel model, which uncovers a nonlinear relationship between cabinet stability and economic performance. Empirical estimates indicate that each cabinet reshuffle lowers annual GDP per capita growth by roughly 1.7 to 2.9 percentage points. The study further detects an inverted U-shaped relationship between ministerial tenure and economic outcomes. The Least Squares Dummy Variable Corrected (LSDVC) estimator reveals that ministers maximize economic growth at an optimal tenure of approximately 51.9 months (4.3 years). Beyond this threshold, moral hazard effects dominate the gains from accumulated experience. These results underscore a pivotal governance dilemma for African presidents: they must weigh the political benefits of cabinet reshuffles against their economic costs. The study thus advances insights into governance dynamics and economic performance in African settings.

Abstract: Numerous studies show that the use of dental guides that rest on the patient's teeth improves the precision of implant placement, but the currently developed procedures and selected materials are still not perfect and could result in deviations from the planned implant position. The impact of the surgeon's hand force on the deformation of dental guides during implant placement has not yet been investigated or documented. In this study the behavior of the 3D guide model is evaluated by FEA analysis to validate the influence of the force of the surgical hand on dental guides due to their application. FEM simulation of deformation and stress was designed for four different types of dental guides that are supported on teeth for different ways of supporting the guide due to the action of the surgeon's manual force (chosen arbitrarily 30 N). The finite element simulation method performed on 5 sets of commonly used biocompatible polymer materials, Stratasys MED610 and VeroGlaze MED620, EOS PA2200, Formlabs FLSGAM01 and Stratasys ULTEM 1010, successfully numerically quantified the deformation of the dental guide caused by the surgeon's arbitrarily manual forces during manipulation. Based on the conducted analyses, guidelines were proposed for improving the design of guides, with an emphasis on optimal selection of supports, stability on the patient's anatomy, and reduction of deformations, thereby increasing the accuracy of implant placement. It was found for all four designs of dental guides that the deflection depends on the size of the arm, i.e. the distance of the support from the point of application of the force. As a result of the study, diagrams were created that serve as guidelines for the design of beam (A and A1) and cantilever (B and B1) versions of dental guides that rely on teeth. Guidelines for enhancing guide design were put out based on the analyses that were carried out, with a focus on the best choice of supports, stability on the patient's anatomy, and minimization of deformations in order to improve implant placement accuracy.

Abstract: The current consensus model of sepsis is that it is a dysregulated host response to infection associated with severe organ dysfunction and failure. In 2023 the author proposed a new model of sepsis in that it was a physiological response and defence to infection that failed or became “dysregulated” particularly if the infection was overwhelming or there was a deficiency of thiamine and/or intracellular glucose to provide ongoing fuel for the immune response and/or mitochondrial production of adenosine triphosphate (ATP).This new model proposed that during sepsis, the immune system received priority access to available glucose, prompting insulin resistance that minimised glucose utilisation by less essential tissues. Concurrently, mitochondrial ATP production via oxidative phosphorylation (OXPHOS) was deprioritised, with the immune system relying on anaerobic glycolysis for ATP generation. This suppression of OXPHOS was only a temporary measure; mitochondrial ATP production had to be resumed for complete recovery. Its persistent suppression could culminate in critical ATP deficits and cell death.This paper reviews the consensus model of sepsis and evidence for the new model.It also reviews glucose, thiamine and insulin metabolism in sepsis and concludes that administering high-dose insulin alongside mild hyperglycaemia and intravenous thiamine—a pyruvate dehydrogenase kinase (PDK) inhibitor—may help restore physiological mitochondrial ATP production when administered during a crucial window in the sepsis process, potentially improving survival outcomes.The thrust of this new model may have been validated by a recent experiment on sepsis in mice that found superior survival following treatment with combined glucose and thiamine compared to antibiotics.

Abstract: Professional practice in clinical and educational contexts frequently operates under a “naive realist” assumption: that professionals and subjects inhabit an identical world, accessible through standardized metrics. This presumes that the professional’s sensory capture constitutes an objective baseline rather than a specific, stabilized interface. We argue that no practitioner—particularly in mental health or education—can legitimately assume that their perceptual field equates to ontological reality. This error drives a systematic structural failure we term epistemic appropriation: the professional’s fundamental inability to conceive that the Other inhabits a legitimately different perceptual reality. Drawing on the Trace & Trajectory Framework (TTF) and Interface Theory of Perception, we provide formal tools to map what critical literature terms “minoritization” and “epistemic injustice.” Epistemic appropriation operates through three complementary manifestations: flattening (the dominant agent’s geometric collapse of the Other’s autonomous identity space into a marginal subset of their own navigational terrain); internalization (the subalternized agent’s coerced construction of self-access routes through dominant-compatible positions); and trajectorial refraction (resistance operations that disrupt the appropriation cycle—counter-exonymy and autonymy exemplify this mechanism). We model this asymmetry endogenously through an Asymmetry Function (gasym) that specifies how differential stabilization dynamics produce navigational inequality: massive institutional recurrence generates configurations whose elevated maintenance costs are collectively distributed and rendered phenomenologically invisible through mimetic naturalization, thermodynamically overwriting fragile autonomous trajectories that bear their full maintenance burden without collective support. Critically, intersubjectivity occurs not through shared ontological space but via transductive coupling—systematic correspondence between distinct navigational interfaces, where asymmetry resides not in the coupling mechanism itself but in the differential Transductive Coupling Costs imposed by exchange protocols and historically saturated semiotic patterns. This formalization distinguishes epistemic appropriation from internalized oppression, provides operational indicators for detecting these dynamics in professional practice, and makes implicit ontological erasure analytically tractable.

Abstract:

As a core component of mechanical equipment, the operational status of bearings directly determines equipment safety, making early fault diagnosis critically important. However, bearing vibration signals are susceptible to substantial noise interference and exhibit both nonlinear and non-stationary characteristics, rendering traditional single-mode diagnostic methods ineffective at extracting fault features. Therefore, this paper proposes a three-channel multimodal fault diagnosis network (M-CNNBiAM) integrated with a convolutional autoencoder (CAE). Based on a convolutional neural network (CNN) architecture, this network employs CAE for signal denoising, utilizes continuous wavelet transform (CWT) to construct time-frequency features, and incorporates dual enhancement modules: convolutional attention (CBAM) and window attention (S-W-MSA).On one hand, it extracts complementary features from the raw vibration signal and the wavelet transform frequency domain signal, fusing them at the channel dimension. On the other hand, it embeds Shifted Window Attention (SW-MSA) and Window Self-Attention (W-MSA) between convolutional layers to capture global-local features. Combined with CBAM to enhance fault location attention, it mitigates the vanishing gradient problem through residual connections, enabling the extraction of frequency domain features. To address the characteristics of one-dimensional time-series signals, a bidirectional gated recurrent unit (BiGRU) is introduced to collaborate with CNN for extracting temporal features. Experiments demonstrate that on the West China University public dataset and self-test dataset, M-CNNBiAM achieves an average diagnostic accuracy of 95.84% under -10dB high-noise conditions, outperforming comparative methods and validating its superior performance in complex noise environments.

Abstract: Background/Objectives: Suicidal ideation and non-suicidal self-injury (NSSI) represent major public health concerns among adolescents, yet developmentally appropriate, school-based interventions remain limited. This study reports findings from an explora-tory analysis of an early cohort of an ongoing randomized controlled trial evaluating Sandplay Therapy with Suicidal Ideation and Self-Injury–Focused Engagement (SPT-SAFE) compared with Treatment as Usual–Risk Management Counseling (TAU-RMC) in a school-based high-risk intervention setting. Methods: Adolescents aged 12–19 years presenting with suicidal ideation and/or NSSI were randomly assigned to SPT-SAFE (n = 31) or TAU-RMC (n = 30). Outcomes of interest were NSSI frequency, assessed using the Functional Assessment of Self-Mutilation (FASM), and suicidal ideation severity, assessed using the Suicidal Ideation Questionnaire–Junior (SIQ-JR). Prespecified baseline-adjusted analyses of covariance (ANCOVA) were conduct-ed as the primary analytic approach. Sensitivity analyses using linear mixed-effects mod-els (LMMs) were performed to examine outcome trajectories over time. Results: In the prespecified baseline-adjusted ANCOVA, suicidal ideation showed a be-tween-group difference favoring SPT-SAFE. For NSSI frequency, the between-group effect also favored SPT-SAFE but was small and did not reach conventional statistical signifi-cance. Sensitivity analyses using LMMs demonstrated directionally consistent patterns, with greater reductions over time observed in the SPT-SAFE group across outcomes. No serious adverse events were reported. Conclusions: Findings from this exploratory early cohort analysis suggest a preliminary and hypothesis-generating signal of benefit associated with SPT-SAFE in a school-based setting, characterized by directionally consistent patterns across complementary analytic approaches. Results should be interpreted as provisional pending completion of recruit-ment, longer-term follow-up, and further evaluation of comparative effectiveness, durabil-ity of treatment effects, and mechanisms of change.

Abstract: Quantum mechanics reveals that physical quantities and informational states are not absolute but relational, depending on the context of interaction between systems. While classical physics already contained relational elements—most clearly in Galilean relativity and Einstein’s relational spacetime—the quantum domain extends relationality to physical properties and facts themselves. In this paper, I develop an info-computational perspective on relational quantum mechanics (RQM), conceiving observers as informational agents embedded within physical processes. Quantum states are understood as constraints on possible interactions rather than intrinsic attributes of isolated systems. I review key relational, perspectival, and information-theoretic approaches—including QBism, perspectival quantum realism, reference-frame–dependent observables, categorical quantum mechanics, and graph-based formalisms—and argue that they converge on a view of physics grounded in relations and information flow. Relational objectivity emerges through inter-agent translation rules rather than observer independence, providing a unified framework for understanding quantum measurement, inter-observer agreement, and physical ontology.

Abstract: We present a consolidated, referee-auditable formulation of the Quantum Information Copy Time(QICT) program. A single localized information-theoretic object—relative entropy, equivalently amodular-energy deficit—is shown to (i) control restricted operational distinguishability via dataprocessing and Pinsker-type inequalities and (ii) coincide with the variational functional enteringentanglement-equilibrium gravitational closure through the exact modular identity D = ∆⟨K⟩−∆S. We separate exact information-theoretic statements from regime-dependent field-theoreticassumptions (local modular Hamiltonians in small causal diamonds) and from microscopic proposals.A reproducible microscopic lattice toy model numerically verifies the operational bounds and themodular identity with embedded figures generated by code. Finally, we include a concrete discreteinformation-field model class formulated by a local gauge-invariant action on a causal cell complexand specify nonperturbative decision criteria under which General Relativity may arise as an infrareduniversality class, without claiming that this emergence is established here.

Abstract: This paper explores the implementation and strategic development of Electronic Records Management Systems (ERMS) across diverse governmental contexts, with attention to both developed and developing nations. Grounded in ISO 15489:2001, the paper examines the core functions of ERMS, including the creation, maintenance, storage, and disposal of digital records. It outlines key implementation strategies, such as policy development, stakeholder engagement, and data migration and conversion. The findings highlight common challenges, such as inadequate infrastructure, limited Internet access, and shortages of skilled personnel. Conversely, the paper emphasizes the benefits of ERMS, including secure information handling, improved organizational efficiency, and enhanced service delivery. This work contributes to the field of information management by providing a practical and comparative overview of ERMS adoption. It identifies critical success factors and offers guidance for policymakers and practitioners aiming to enhance record management in the digital age.

Abstract: As intelligent autonomous systems (IAS) continue to assume increas-ingly central roles in safety- and mission-critical domains such as transportation,healthcare, finance, and infrastructure management, humans are becoming una-ble to monitor or intervene in real time. This shift is driven by the speed, data-processing capacity, and adaptivity of IAS. To manage this complexity, a newparadigm is emerging: IAS controlling and monitoring other IAS, a developmentthat introduces at the same time practical efficiency and profound practical andethical challenges.This article explores the multi-layered delegation of responsibilities within IASecosystems, where decisions influencing human lives and well-being are madewith minimal human intervention. One often-overlooked consequence of this del-egation is the capacity of AI systems to shape and create new human habits,whether through personalized persuasion, behavioral feedback loops, or autono-mous decision enforcement. As humans increasingly adapt their behaviors to ma-chine-optimized environments, questions arise about autonomy, agency, and re-sponsibility for resulting behavior changes.Drawing on insights from recent research on responsibility delegation in IAS andon AI-driven habit formation, the article critically examines how responsibilityshould be distributed across human actors, autonomous systems, and institutions.Framed within the principles of Digital Humanism, I argue for a value-sensitivegovernance model that ensures transparency, explainability and human oversighteven in complex IAS-to-IAS control scenarios.I propose a normative framework for responsibility attribution that accounts forboth the technical architecture of IAS networks and the behavioral effects thesesystems have on human users. The article concludes by addressing the ethicalrisks of diminished human agency, manipulation through behavioral design, andthe need for institutional mechanisms that align IAS operations with fundamentalhuman values.

Abstract: This paper addresses the mind-body problem by highlighting the concept of continuity. Building on philosophical definitions, it introduces an "ontology of continuity" thesis to bridge the mental and the physical. Based on the thesis, the paper introduces a “neuro-subjective interactive (NSI)” model, which incorporates empirical studies from brain science. The model suggests that subjective experiences and neural activities are inter-dependent. Instead of separating between human mentality and its physical base, the paper posits that our mentality is constituted by both physical (neural) and non-physical (subjective) elements. This approach addresses two major challenges in the mind-body problem: causal overdetermination and physical causal closure.

Abstract: Drag reduction forms a key area of focus in aerodynamics with a significant emphasis on delaying the laminar to turbulent transition of boundary layers over the wing of aircraft. There is enough evidence to suggest that achieving such transition delays is particularly challenging for backward swept wings with large leading edge sweep angles, which give rise to crossflow and attachment line instabilities, in addition to the Tollmien-Schlichting waves. The sustenance of extended laminar flow regions at high sweep angles has been demonstrated in recent studies, by designing airfoils with specially curated leading edge profiles, which generate pressure distributions that can suppress crossflow. Such airfoils are called Crossflow Attenuating Natural Laminar Flow (CATNLF) airfoils. However, the design of such airfoils is presently restricted to inverse methodologies due to the inability of the conventional geometry parameterization techniques in representing the specialized leading edge profiles of CATNLF airfoils. The aim of this study is to illustrate that a parametric representation of CATNLF airfoils can be realized using Bezier curves, thereby enabling their forward multi-point design using gradient-free Bayesian optimization. The developed design framework in terms of geometry parameterization and optimization formulation is able to deliver airfoils that can sustain natural laminar flow up to around 50% chord length on the upper surface, with a leading edge sweep angle greater than 27 degrees at a Mach number of 0.78 and a Reynolds number of 20 million within a range of lift coefficients Cl = 0.5 ± 0.1, making them a suitable design choice for a medium-range transport aircraft.

Abstract: Rapid economic growth does not necessarily translate into better perceived urban health. Using the 2024 nationwide Urban Physical Examination (UPE) resident survey in China, this study assesses how city economic level relates to perceived urban health, proxied by city-level overall satisfaction. The survey was conducted in April–June 2024 in the main urban districts of 47 cities, yielding 692,800 responses and 499,500 valid questionnaires. We aggregate satisfaction to the city level, match it with GDP and key city characteristics, and estimate the GDP–satisfaction association using restricted cubic splines (RCS) to test for potential non-linearity. Across unadjusted and covariate-adjusted models (accounting for population scale and density, industrial structure, fiscal capacity, and regional effects), results show a robust positive association between economic level and satisfaction, while nested-model tests provide no evidence that spline terms improve fit over a linear specification within the observed GDP range. Substantial dispersion around the fitted curve indicates that GDP is an enabling capacity rather than a sufficient condition, pointing to cross-city differences in how effectively resources are converted into lived urban quality. We propose using GDP-adjusted satisfaction benchmarking within the UPE cycle to identify underperforming cities and prioritize targeted governance and renewal actions.