Abstract: This study examines the aerodynamic behaviour of thin perforated plates through a combined experimental and numerical methodology integrating wind-tunnel measurements, fully resolved CFD of the test section, and computationally efficient periodic ``modulus'' simulations. The objective is to provide reliable and transferable drag coefficients for porous plates employed in façade engineering and flow-control applications.The three standard approaches for estimating aerodynamic drag (force balance, total-pressure drop, and static-pressure difference across the plate) are systematically compared under imposed flow-rate conditions. Although often treated as equivalent, the methods yield non-coincident results. High-resolution CFD demonstrates that the static-pressure field on the windward face of the plate is intrinsically non-uniform, leading to a systematic overestimation of drag when pointwise static-pressure measurements are used. This motivates the introduction of a physically based correction factor, γ ≈ 5%, which is experimentally validated and enables static-pressure estimates to be aligned with force-balance data.Once validated, simulations in cyclic ``modulus'' configuration (where only the smallest repeating unit of the perforated plate is simulated) accurately reproduce the global aerodynamic response of the plates at a greatly reduced computational cost, enabling extensive parametric analyses. Results show that porosity is the dominant parameter governing drag, whereas the hole pattern mainly affects local flow structures with limited influence on the integrated force.

Abstract: Is an ant colony conscious? What about a group of people talking, a cloud-hosted language model, or even a galaxy? Can a conscious mind only get so big? Does consciousness depend only on what is computed, or when and where? I see two possibilities affecting the answers to these questions. I name them Chord and Arpeggio, and formalize the distinction mathematically. If the ingredients of a subjective experience must be simultaneously true at one objective instant and causally exchange influence within a time window θ, then the system diameter D satisfies D ≤ κvθ, where v is the signal speed ceiling and κ depends on exchange architecture. I call this requirement Chord, because it is like a musical chord whose notes sound together. The alternative is Arpeggio. It asks only that each ingredient occur somewhere in the window. I prove that Arpeggio is strictly weaker than Chord, and that architectures with limited concurrency can satisfy Arpeggio while structurally forbidding Chord. I argue for Chord on formal, neuroscientific, and architectural grounds. A mechanistic model confirms a fragmentation transition at the theoretical threshold. I examine primate corpus callosum data to estimate empirical lower bounds on θ. I provide case studies showing that under Chord, ant colonies and human populations are ruled out as single conscious entities, cloud-hosted AI is constrained by co-instantiation rather than diameter, and brain-computer interface hybrids face latency-dependent limits. A mind can only get so big. Arpeggio is far more permissive, implying consciousness seemingly everywhere.

Abstract:

Marine exopolysaccharides (EPS) are emerging as sustainable bioactive polymers for biomedical hydrogels. Here, we report hydrogels from sulfated EPS produced by Porphyridium cruentum and ionically crosslinked with Ca²⁺, Ce³⁺, or Cu²⁺ to generate tunable networks for wound-healing applications. Rheological analysis showed that viscoelastic behavior was primarily governed by cation nature and accessible binding-site density, with diminishing gains above 2.5 wt% EPS and limited benefit beyond 10 wt% crosslinker. Ce³⁺ produced the most solid-like gel, Ca²⁺ yielded more thixotropic networks, and Cu²⁺ promoted rapid, heterogeneous crosslinking consistent with fast surface complexation. These network signatures translated into distinct in vitro performances. Cation selection tuned antibacterial activity against Staphylococcus aureus and Escherichia coli, with Cu²⁺ achieving rapid bactericidal effects and Ce³⁺ enabling an 8-log reduction after 24 h. Antioxidant capacity was assay-dependent (ABTS vs DPPH), reflecting combined EPS radical-quenching and metal-associated redox contributions. Conditioned-media assays using human dermal fibroblasts and keratinocytes indicated the most favorable cytocompatibility balance for Ce³⁺-crosslinked gels, whereas Cu²⁺ gels were limited by cytotoxicity. Macrophage cytokine readouts (TNF-α, IL-6) further supported formulation-dependent immunobiological activity. This work establishes microalgal EPS as a versatile polymer platform and links ionic crosslinking chemistry to rheological control and multifunctional biomedical performance.

Abstract: Cavitation in rotating hydraulic machinery -- such as industrial pumps and hydropower turbines -- can cause blade and casing erosion, excessive vibration, noise and efficiency loss, posing significant operational and economic risks across industrial sectors. Reliable and scalable monitoring strategies are therefore essential, particularly under variable operating conditions in real-world environments. Recent advances in machine learning (ML) and deep learning (DL) have enabled data-driven approaches for cavitation detection based on operational sensor signals, yet a structured synthesis of these developments is lacking. This scoping review systematically analyzes measurement-based ML and DL approaches for cavitation monitoring, with the aim of identifying key trends, challenges and future research directions. Following PRISMA-ScR and JBI guidelines, 52 peer-reviewed studies published between 1996 and 2025 were evaluated, covering laboratory and field investigations across pumps and turbines and a wide range of model architectures. The analysis reveals that most studies are laboratory-based (∼ 80%), focus on pumps (∼ 70%) and rely on single-machine datasets (> 80%), limiting generalization across machines and operating conditions. Classical ML approaches remain relevant due to interpretability and robustness with limited data, while DL enables end-to-end learning from raw or time-frequency transformed signals, frequently achieving diagnostic accuracy above 95%. Hybrid frameworks combining DL-based feature extraction with classical classifiers are increasingly adopted. Key limitations across the literature include domain shifts between laboratory and field data, scarce or inconsistent labeling and a predominant focus on categorical cavitation severity levels.

Abstract: Hydrodynamic cavitation is attracting increasing interest in food processing as a non-thermal approach for preserving product quality and supporting the recovery of valuable bioactive compounds. Conventional Venturi devices are usually designed for fixed operating conditions, whereas real process streams may vary in temperature, viscosity, and gas or solid content. This can make it difficult to maintain stable and effective operating conditions when a fixed geometry is used. In this work, an adjustable circular Venturi is presented as a simple conceptual device for hydrodynamic cavitation in food applications. The external body and pipeline connections remain unchanged, while the throat section can be adjusted to adapt the device to different process requirements. In this sense, the proposed concept may also serve as an adjustable platform for exploring different operating conditions and identifying suitable throat configurations for specific food matrices and process targets. Once identified, such conditions may support the definition of a dedicated final Venturi configuration for the intended application. The proposed concept may be of interest for applications such as green extraction, food by-product valorization, and mild processing strategies aimed at preserving or enhancing bioactive compounds. This study is presented as a conceptual design contribution for food applications.

Abstract: Periodontitis is a chronic inflammatory disease driven by microbial dysbiosis and an exacerbated host immune response, leading to progressive periodontal tissue breakdown and contributing to systemic inflammation. Although scaling and root planing remains the standard treatment, its capacity to fully restore immune balance and host–microbiota homeostasis is limited. In this context, probiotics have emerged as promising adjunctive strategies capable of modulating immunological and metabolic pathways involved in disease progression. This narrative review aimed to evaluate current evidence regarding the use of probiotics in periodontal therapy. The review followed the Scale for the Assessment of Narrative Review Articles (SANRA) guidelines. A literature search was conducted in MEDLINE via PubMed for manuscripts indexed up to January 2026 using MeSH-based terms related to periodontitis and probiotics. Evidence from preclinical and clinical studies suggests that probiotics may reduce alveolar bone loss and periodontal inflammation by downregulating proinflammatory mediators, enhancing anti-inflammatory cytokine production, strengthening epithelial barrier function, and modulating innate and adaptive immune responses. Additionally, probiotics may exert systemic effects through interactions with the gut microbiota, potentially improving metabolic regulation and reducing systemic inflammation. Overall, current evidence supports probiotics as biologically plausible adjuncts to periodontal therapy.

Abstract: The management of today’s optical networks is highly dependent on the correct estimation of Quality of Transmission (QoT). The current analytical approach requires exact physical values, which are often not available, resulting in inefficient management of the network. This paper proposes an Adaptive Machine Learning Framework that aims to address the analytical approach’s limitations using a new and innovative data-driven approach. The proposed framework combines linklevel embeddings with an Artificial Neural Network (ANN) to process the unique sequence of fiber links in a lightpath, focusing on the fine-grained details of the sequence that are normally overlooked by the current analytical approach. Through dynamic learning from the sequence data, the framework provides highly accurate signal quality estimates. These estimates enable intelligent and automated modulation format choices, greatly enhancing spectral efficiency and minimizing disconnections. This highly scalable solution is developed in Python and TensorFlow and is best suited for dynamic resource allocation and futureoriented network planning.

Abstract: Research on adaptive strategies in extreme environments is crucial for understanding the resilience of human survival wisdom. This study integrates multidisciplinary evidence from archaeology, zooarchaeology, archaeobotany, isotopic, and geochemical analysis to reassess the role of fishing, hunting, and gathering economies in prehistoric arid Xinjiang, northwestern China. Our findings reveal that, spatially, fishing concentrated in the Lop Nur region of the Tarim Basin, with potential activities extending to the surrounding river basins across the Altai, Tianshan, Pamir, and Kunlun mountains; hunting was more developed in Northern Xinjiang (focusing on deer and bovids) while practiced on a smaller scale in Southern Xinjiang (targeting hares); gathering also exhibited north–south divergence in plant utilization. Temporally, these economies declined from a dominant Paleolithic strategy to a supplementary role in the Bronze and Early Iron Ages. However, resilient local adaptations persisted—notably at Lop Nur (fishing), Xiaxingguang cemetery (specialized hare hunting), and the Eastern Tianshan region (high-proportion gathering). Beyond subsistence, these practices were deeply embedded in spiritual life, reflected in totemic symbols and shamanic ritual paraphernalia.This study re-evaluates prehistoric extractive economies, providing critical insights into human adaptation strategies in arid to semi-arid environments.

Abstract: Background/Objectives: Lower-limb total joint arthroplasty (TJA) has been associated with neuroprotective effects, including reduced incidence of dementia and Parkinson disease. Whether these effects are mediated by restored ambulation (specific to lower-limb surgery) or by the systemic anti-inflammatory consequences of arthroplasty (shared by all joint replacement procedures) remains undetermined. We used total shoulder replacement (TSR) as a negative control comparator to interrogate this mechanistic question. Methods: Using the TriNetX US Collaborative Network (114 million patients), we constructed propensity score-matched cohorts comparing TSR patients to total knee arthroplasty (TKA) patients (66,038 per group) following a 730-day lag period. Five pre-specified outcomes were tracked: incident dementia, sarcopenia, Parkinson disease, cataract (active positive control), and elevated C-reactive protein (CRP). Matching balanced 18 demographic and comorbidity covariates. Kaplan–Meier survival analysis with log-rank testing and Cox proportional hazards regression were performed. Results: After propensity matching, TSR and TKA groups were balanced on all covariates (standardized mean differences < 0.10 for all 18 variables). Compared to TKA, TSR patients showed no significant reduction in incident dementia (HR = 1.63, 95% CI: 0.75–3.55, p = 0.217) or sarcopenia (HR = 1.32, 95% CI: 0.72–2.45, p = 0.369). Notably, TSR patients had significantly higher rates of incident Parkinson disease (HR = 1.24, 95% CI: 1.08–1.41, p = 0.002) and more frequent CRP elevation (HR = 1.13, 95% CI: 1.06–1.21, p < 0.001) than TKA patients. The cataract control outcome did not differ between groups (HR = 0.98, 95% CI: 0.89–1.08, p = 0.698). Conclusions: TSR does not replicate the neuroprotective effects associated with lower-limb TJA, and is associated with greater inflammatory burden and higher Parkinson disease incidence than TKA after careful propensity matching. These findings support the hypothesis that restored ambulation—rather than surgical anti-inflammatory effects alone—is the primary mediator of neuroprotection following lower-limb arthroplasty, with important implications for understanding the biology of exercise-dependent neuroprotection.

Abstract: Mathematics, as actually practiced, operates as a federated system: practitioners work within autonomous domain-specific axiomatizations (geometry, algebra, analysis) and construct explicit bridges only when cross-domain reasoning is required. This organization is not accidental; it is a structural adaptation that safeguards local decidability and algorithmic efficiency.Yet the dominant foundational narrative still operates on the Compiler Myth—the belief that all mathematics must theoretically compile down into ZFC set theory to achieve rigor. We argue that this monolithic reductionism confuses representational universality with logical priority. Embedding a decidable (tame) domain into an undecidable (wild) one does not clarify foundations; it imposes a crippling epistemic overhead. It buries efficient, domain-specific decision procedures under general proof search and destroys the native structural immunities of the object.We introduce the Decidability Threshold — a litmus test based on Negation, Representability, and Discrete Unboundedness — to explain why mathematicians instinctively isolate tame domains from wild ones. Finally, we distinguish the Mathematician (builder of formal systems) from the Scientist (consumer modeling reality). We argue that federalism, through explicit bridges and domain autonomy, is not a failure of unification, but the primary safeguard preventing the scientist from inadvertently importing wild, undecidable paradoxes into physical theories.

Abstract: Digital Twin (DT) systems combining physics-based simulation with hardware execution are critical for Industry 4.0 manufacturing, yet proprietary software solutions remain expensive and platform-dependent. This work addresses three technical challenges: maintaining geometric and kinematic fidelity across CAD-to-simulation conversion pipelines, synchronizing dual physics engines (Unity and ROS middleware) under hardware latency constraints, and optimizing motion planning while preserving trajectory quality and interactive responsiveness. We developed an integrated framework for a 7‑Degree of Freedom manipulator using CAD modeling, URDF/SRDF semantic representation, and bidirectional Unity-ROS (Robot Operating System) communication via WebSocket connectors. Motion planning uses RRTConnect from OMPL with collision-aware optimization through the Flexible Collision Library. Validation across 12 manipulation trials demonstrated positional synchronization accuracy of ±2.0 degrees, motion planning performance of 0.064 ± 0.020 seconds. Latency analysis reveals that hardware execution to be the dominant system bottleneck, significantly exceeding network communication delays. The system achieves performance metrics comparable to proprietary industrial solutions. This work establishes a replicable, cost-effective Industry 4.0 framework, demonstrating that modern game engine technology combined with open-source robotics middleware can deliver DT systems matching proprietary solutions. The architecture and validated implementation enable adaptation to alternative robotic platforms and support broader adoption of simulation-validated automation in manufacturing contexts.

Abstract: Public agent ecosystems are emerging as a new object of study in NLP: settings in which language models not only generate text but also act, coordinate, authenticate, exchange reusable capabilities, and leave durable public traces. Using the OpenClaw--Moltbook ecosystem as a strategically revealing case, we survey a curated corpus of 38 ecosystem-specific papers and reports available as of 10-03-2026, together with official platform materials and adjacent survey literature. We provide a case-centered, NLP-centered survey of a public agent ecosystem in the wild. We argue that this case is best understood as language infrastructure: linguistic artifacts are executable, persistent, public, portable, and increasingly governance-bearing. We introduce GATE --- Grounding, Action, Transfer, and Exchange --- to organize what language does in public agent ecosystems, and pair it with AERO --- Authority, Enablement, Reach, and Orchestration --- to track how language acquires delegated operational force. Across the corpus, the main methodological bottleneck is weak triangulation across trajectories, discourse, portable artifacts, and grounding signals. That bottleneck yields four recurring fault lines: instruction is mistaken for authority, visible agent speech is mistaken for autonomous speakerhood, public claims outrun verification, and local control is mistaken for lower risk. We conclude with an NLP agenda centered on executable pragmatics, delegated-agent discourse analysis, provenance-aware evaluation, privacy-preserving agent NLP, multilingual public-agent research, and autonomy-sensitive benchmarks. We will release all artifacts once permitted.

Abstract: We study a pharmaceutical scheduling problem with hybrid batch-continuous manufacturing process in a distributed supply chain. The supply chain consists of heterogeneous plants and one distribution center. Each plant adopts an unrelated permutation flow shop layout consisting of a hybrid batch-continuous production line. Each pharmaceutical order is split and produced in multi-production sites located in various regions. The pharmaceutical medicines manufactured by the production sites are directly shipped to a distribution center To minimize the makespan, we formulate the addressed scheduling problem as a mathematical model. To solve this model, we propose four metaheuristics variants by applying two population-based metaheuristics to two distinct solution structures. We compare the proposed metaheuristics to evaluate their performance in the numerical experiments. Additionally, we present managerial insights through sensitivity analysis.

Abstract:

Background/Objectives: The escalating crisis of antibiotic resistance and the inherent limitations of conventional antibiotics necessitate the development of innovative therapeutic strategies. Targeted drug delivery (TDD) offers a powerful approach to enhance efficacy, minimize systemic toxicity, and circumvent bacterial resistance. This systematic review aims to evaluate the potential of unique bacterial transport systems (BTSs) and surface specific receptors as platforms for TDD via the "Trojan Horse" strategy (THS). Methods: A comprehensive literature review was conducted, focusing on studies that investigated the specificity and mechanisms of BTSs responsible for the uptake of essential metabolites. This includes an analysis of transport systems for siderophores, bacteria-specific sugars, cell wall components, D-amino acids, and vitamins. We assessed preclinical and clinical examples of drug conjugates utilizing these pathways, as well as emerging platforms such as bacteriophage-derived proteins, antibody-antibiotic conjugates, and bacterial extracellular vesicles (EVs). Results: BTSs demonstrate high specificity for their cognate substrates, providing effective molecular gateways for drug conjugate import. The siderophore-cephalosporin conjugate cefiderocol represents a clinically validated example, having received FDA approval. Preclinical studies further reveal that conjugates utilizing sugars (e.g., maltose, trehalose) and vitamins (e.g., B12) can significantly enhance antibiotic uptake and activity against both Gram-positive and Gram-negative pathogens, including drug-resistant strains. Emerging platforms like bacteriophage endolysins and engineered EVs show promise for overcoming biological barriers such as bacterial outer membranes and intracellular host niches. Conclusions: The THS leveraging BTSs represents a clinically viable and promising avenue for next-generation antibacterial therapies. While significant progress has been made, including regulatory approval of cefiderocol, further research is critically needed to identify novel BTSs, optimize drug-linker chemistry, improve the pharmacokinetics and biosafety of conjugates, and translate these innovative platforms into effective treatments for drug-resistant infections.

Abstract: Maximum oxygen consumption (VO2max) is the ability to absorb, transport, and use oxygen in the body to produce useful energy for muscle activation in a unit of time. For years, devices have been developed to estimate physical performance, including VO2max. In view of the above, it can be considered that the use of artificial intelligence can facilitate interpretation and even generate estimates based on available data. In this regard, this study aims to review the use of artificial intelligence in the assessment of maximum oxygen consumption. A scoping review was conducted in accordance with the following stages: (i) identification of the research question: What would be the use of AI to predict VO2? (ii) identification of relevant studies: searching academic databases and AI search engines; (iii) selection of studies: the PRISMA ScR protocol was applied, selecting 50 studies; (iv) graphing the data in the results (v): finding studies published since 2009 with a higher publication rate in countries in the Americas and Asia; it is concluded that the use of deep learning fed with validated algorithms allows for a more accurate estimation of VO2max and that its evaluation requires the use of explainable AI training, starting with the linear regressions available in the literature and continuing with decision trees, to predict performance and offer a classification of it.

Abstract: Raman Spectroscopy is non-destructive, label free analytical technique that can probe the biochemical alterations in tissues and cell. Raman Spectroscopy, being sensitive to biochemical perturbations, can potentially provide early and real-time identification of changes proceeding morphological changes, allowing early diagnosis as well as diseases monitoring. Recent research has demonstrated its broad utility across diverse clinical domains, including cancers, neurological conditions and infections. Raman spectroscopy combined with machine learning algorithms allows rapid assessment and automated pipelines and can act as a clinical adjunct, enhanced by integrating tools like principal component analysis (PCA), linear discriminant analysis (LDA), random forests, and deep learning architectures. These models allow interpretation of complex spectra, and decipher meaningful biomarkers in heterogeneous clinical samples. This review highlights the earliest and recent progress in Raman based non-destructive diagnosis underscoring advances in cancer diagnosis and challenges faced in clinical settings.

Abstract: Fatigue is a leading contributor to maritime accidents, yet recreational sailors lack the regulatory frameworks and fatigue management tools available to commercial mariners. Peer-reviewed research published in Nature demonstrates that after 17 hours of sustained wakefulness, cognitive performance degrades to a level equivalent to a blood alcohol concentration of 0.05% — the legal driving limit in most countries (Dawson & Reid, 1997). After 24 hours, this rises to 0.10%, well past the threshold for legal intoxication. These findings have been independently replicated (Williamson & Feyer, 2000) and confirmed in field studies aboard racing yachts (Hurdiel et al., 2014). This paper synthesises more than three decades of peer-reviewed research spanning chronobiology, sleep medicine, occupational health, and maritime safety into a biomathematical fatigue model calibrated specifically for pleasure boat passage-making. The model integrates sleep-wake homeostasis, circadian rhythm modulation, sleep fragmentation effects, environmental sleep degradation from sea state, and cumulative multi-day sleep debt into a single framework that outputs impairment as a BAC (blood alcohol concentration) equivalence — an intuitive metric that any sailor can understand. Critically, the model is not merely theoretical. It has been implemented as a freely available, open-access passage fatigue calculator for mobile and web platforms, making it accessible to the widest possible population of recreational mariners. The application faithfully reproduces every formula, constant, and coefficient described in this paper, allowing sailors to simulate any passage plan — varying crew size, watch schedule, departure time, pre-departure sleep, and sea state — and see the predicted fatigue trajectory hour by hour. The purpose is to bridge the gap between laboratory science and practical seamanship: to give pleasure boat crews the same evidence-based fatigue awareness that professional mariners receive through regulation.

Abstract: Multilocular intratesticular cysts are uncommon benign lesions. We report a case associated with testicular microlithiasis in an 85-year-old man presenting with painless enlargement of the left scrotum. Ultrasonography revealed a multilocular cystic lesion with a 3.0-cm main cyst and several adjacent smaller cysts showing posterior acoustic enhancement without mural irregularity or solid components. Bilateral microlithiasis and small epididymal cysts were also detected, and serum tumor markers were normal. The lesions remained stable during 36 months of follow-up. Recognition of the characteristic ultrasonographic features of benign intratesticular cysts is important to avoid unnecessary surgical intervention.

Abstract: Background/Objectives: Malnutrition is highly prevalent in patients with advanced colorectal cancer (CRC) and is associated with poorer treatment tolerance, reduced quality of life, and worse clinical outcomes. Chemotherapy regimens incorporating irinotecan are commonly used in later treatment lines but present a distinct toxicity profile characterized mainly by gastrointestinal symptoms and functional decline, which may negatively affect nutritional status. This study aimed to evaluate the impact of irinotecan-containing chemotherapy regimens on nutritional status and functional capacity in patients with advanced CRC using the Patient-Generated Subjective Global Assessment (PG-SGA). Methods: A cross-sectional observational study was conducted in 91 adult patients with histologically confirmed stage III–IV colorectal adenocarcinoma undergoing active systemic chemotherapy between November 2023 and June 2024. Demographic and clinical variables, treatment regimens, and exposure to irinotecan were recorded. Nutritional status was assessed using the PG-SGA, and body composition was evaluated by multifrequency bioelectrical impedance analysis. Associations between irinotecan exposure and nutritional, functional, and symptom-related variables were analyzed using χ² or Fisher’s exact tests. Results: Thirteen patients (14.3%) received irinotecan-containing regimens, most frequently combined with FOLFOX. Patients exposed to irinotecan presented higher PG-SGA scores, indicating a greater nutritional burden. Although most gastrointestinal symptoms did not reach statistical significance, early satiety (p = 0.041) and functional deterioration (p = 0.039) were significantly associated with irinotecan administration, while nausea (p = 0.089) and vomiting (p = 0.087) showed trends toward significance. The subgroup treated with FOLFOX–irinotecan also demonstrated a higher frequency of functional impairment compared with those receiving FOLFOX alone. Conclusions: Irinotecan-containing chemotherapy regimens in advanced CRC are associated with a distinct pattern of nutritional deterioration primarily driven by functional decline and gastrointestinal symptoms affecting food intake. Systematic nutritional assessment using validated tools such as the PG-SGA may allow early identification of vulnerable patients and support the implementation of timely multimodal interventions aimed at improving treatment tolerance and clinical outcomes.

Abstract: Accurate estimation of the mean-reversion speed $\alpha$ in the AR(1) process $X_{t+1} = (1-\alpha)X_t + \varepsilon_t$ is central to energy-commodity modelling. Classical estimators such as GARCH, jump-diffusion, and regime-switching produce model-conditioned estimates by embedding $\alpha$ within distributional assumptions, so that different model choices yield different $\hat{\alpha}$ values from the same series without a principled criterion to adjudicate. We propose a distribution-free estimator based on a Temporal Convolutional Network (TCN) trained on synthetic AR(1) series with Sinh-ArcSinh innovations of varying tail weight and asymmetry. The SAS family serves as a training vehicle---not a distributional hypothesis---chosen for its ability to span innovation profiles from near-Gaussian to strongly leptokurtic and skewed through its tail-weight and asymmetry parameters. Because the autocorrelation structure $\rho_k = (1-\alpha)^k$ is invariant to the marginal innovation distribution (Yule-Walker invariance), the TCN learns to extract $\alpha$ from temporal dependence alone, independently of distributional assumptions. On held-out test series the estimator outperforms all three classical estimators across the training innovation kurtosis range, with the advantage growing monotonically with non-Gaussianity. A robustness analysis on three out-of-distribution innovation families confirms stable or improved performance well beyond the training boundary. The distribution-free $\hat{\alpha}$ enables a two-stage pipeline in which $\alpha$ and the innovation distribution are characterised independently---a decoupling structurally impossible in classical likelihood-based approaches. Once trained, the TCN acts as a universal mean-reversion estimator applicable to any price series without re-fitting. Applied to four energy markets---Italian natural gas (PSV price), Italian electricity (PUN price), US Henry Hub, and US PJM West Hub---spanning log-return kurtosis from near-Gaussian to strongly heavy-tailed, the TCN yields robust, model-free estimates not conditioned on any distributional hypothesis.