Abstract: Publicly available data and information relevant and supportive to the 26 August 2026 Langtang Lirung disaster cascade are dispersed across repositories, data portals and other online sources, complicating their identification, assessment and reuse. This study presents a living catalogue (https://evelpidou.nkua.cloud/Nepal-Flood-2026-Catalogue/) of these resources and analyses its first assessed release, archived in Zenodo (https://doi.org/10.5281/zenodo.23034746). The search and charting process, reported in accordance with PRISMA-ScR, comprised 74 documented searches through 24 September 2026, followed by the assessment and reviewing that completed on 28 September. The resulting inventory contains 141 resources and 5,883 assets across eight thematic domains. Structured records preserve provenance, geographic and temporal descriptions, access conditions, rights and source limitations, while the WebGIS (https://evelpidou.nkua.cloud/Nepal-Flood-2026/) enables geographic exploration where spatial information is available. Beyond this initial release, planned periodic searches, event-triggered checks, community nominations and versioned updates provide for continued expansion. The contribution is a traceable inventory connecting pre-event baselines, event observations and subsequent analyses to their supporting material, with documented searches and decisions providing a basis for continued maintenance. The catalogue therefore reduces the time and effort required to identify and access relevant data and information, while lowering the risk of overlooking resources that could inform future investigations of the event. Unlike conventional scoping reviews centred on mapping existing studies or datasets, this approach proposes a new type that systematically identifies and evaluates the data and resources available to support future research.

Abstract: Preferential loss of visceral adipose tissue is clinically desirable because of its strong association with cardiometabolic risk. Hall and Hallgreen (2008) showed that VAT loss is allometrically related to total fat mass loss, with a single dimensionless exponent k ≈ 1.3 describing data from 37 studies and more than 1400 individuals. This relationship is the correct null hypothesis for any claim of selective VAT targeting. The present commentary accepts that foundation and asks whether a mass-balance perspective can both explain it and indicate how to move beyond it. The decisive distinction is between the allometric exponent – an emergent property of the system – and the compartment-specific clearance coefficient – a physiological parameter. Within the mass balance model, VAT's higher clearance coefficient produces the observed preferential early loss under a Torricelli-type relationship between mass and clearance rate. Because the coefficient is modifiable, interventions that alter carbon and lipid mass fluxes may improve the VAT-to-total-fat loss ratio above the allometric baseline. The framework yields six testable predictions, identifies the condition under which the coefficient becomes resistant to change, and provides practical priorities for maximizing visceral fat reduction while preserving lean mass.

Abstract: Background: Hyperprolactinemia, characterized by serum prolactin levels >20 ng/mL, is a rare comorbidity in diabetes mellitus, potentially exacerbating insulin resistance and glycemic dysregulation via JAK/STAT signaling. This case series analysis examines five rare cases to elucidate clinical patterns, molecular mechanisms, and therapeutic outcomes, aiming to inform precision endocrine management. Methods: Five cases were curated from PubMed and Scopus (1999–2025), focusing on hyperprolactinemia in diabetes (type 1, type 2, or insipidus). Data on demographics, symptoms, laboratory findings (prolactin, insulin, C-peptide), imaging (MRI), and treatment outcomes were extracted. Results: The cohort included 3 females and 2 males (mean age: 26.2 years, range: 7–33). Prolactin levels ranged from 145–3,177 ng/mL, with 2/5 cases (40%) having type 1 diabetes, 1/5 (20%) type 2, 1/5 (20%) diabetes insipidus, and 1/5 (20%) factitious hypoglycemia. Presentations included hypogonadism (3/5, 60%), galactorrhea (2/5, 40%), and hypoglycemia (2/5, 40%). MRI revealed microadenomas (2/5, 40%), a macroadenoma (1/5, 20%), or normal pituitary (2/5, 40%). Cabergoline or medication withdrawal normalized prolactin in 4/5 (80%), reducing levels (e.g., 3,177 to 321 ng/mL in one case) and improving glycemic control (fewer hypoglycemic episodes) in 3/5 (60%). Prolactin activates JAK2/STAT5, inducing SOCS3 to inhibit IRS-1/PI3K/Akt, impairing glucose uptake. Conclusion: Hyperprolactinemia, whether tumoral or non-tumoral, disrupts metabolic homeostasis in diabetes. Routine prolactin screening is recommended for patients with reproductive or glycemic anomalies, emphasizing targeted therapies like dopamine agonists to optimize outcomes.

Abstract: Let B \( \subset\ \)R² be the unit disk and let \( A_\beta \) be the positive Robin Laplacian associated with \( a_\beta(u,v)=\int_B \nabla u\cdot\nabla v\,dx+\beta\int_{\partial B}uv\,dS, \qquad \beta>0. \) For \(0<s\le 1\) we study the spectral fractional Lane–Emden problem \( A_\beta^s u=u^p,\qquad u>0\quad\hbox{in }B, \) where the Robin boundary interaction is encoded in the spectral definition of \( A_\beta^s \). The case \( s=1 \) is the classical planar Robin Lane–Emden equation. We first establish variational existence of positive solutions, and of positive radial solutions, in the fractional Sobolev-subcritical range \( p+1<2/(1-s) \). Positivity follows from the positivity-improving Robin heat semigroup and the semigroup formula for \(A_\beta^{-s}\). We then use the classical problem as a nondegenerate endpoint and develop a perturbation theory for \( s\uparrow1 \). In particular, the classical positive radial branch continues to a locally unique fractional radial sheet \( (s,\beta)\mapsto U_{s,\beta} \). The principal result concerns symmetry breaking. For each \( p\ge 8 \), let \(\beta_{p,-}\) and \(\beta_{p,+}\) denote the two distinguished simple first-angular symmetry-breaking parameters of the classical Robin problem established in the published local theory. We prove that, for \(s\) sufficiently close to \( 1 \), these two points persist as two \(C^1\) curves \(\beta_{p,\pm}(s)\) with \(\beta_{p,\pm}(s)\to\beta_{p,\pm}\) as \( s\uparrow1 \). At each point \( (s,\beta_{p,\pm}(s)) \), the fractional linearization has a two-dimensional first-angular kernel in the full space and a one-dimensional kernel in a reflection-invariant subspace. A self-adjoint conjugation of the linearized equation and spectral perturbation theory show that the corresponding simple eigenvalue of the bounded conjugated linearization crosses zero transversally. The Crandall–Rabinowitz theorem then yields local branches of positive nonradial solutions. Hence, for every \( p\ge8 \) and every fractional order sufficiently close to one, uniqueness among all positive solutions fails along parameters converging to each of the two fractional bifurcation curves, while the radial state remains locally unique within the radial class. We also give the extension formulation, the exact disk-scaling law, and explain why a genuinely variable-order model requires a different operator framework.

Abstract: Generalizing a distribution can be achieved by inserting one or more new parameters to the baseline distribution. The author deliberates this methodology to build, Weibull Median Based Unit Rayleigh {Y} family of distributions. The author applies the approach of the T-X{Y} family to add more than one parameter to the baseline distribution Median Based Unit Rayleigh (MBUR) defined on the unit interval with one parameter b > 0. The generator distribution is the Weibull distribution with the two parameters, c & ξ > 0, the shape and scale parameters respectively as will be explained in the paper. The author consumed different link distributions or Y distributions in the form of the quantile function of this link distribution to link the baseline distribution, the MBUR, to the generator distribution, the Weibull distribution. The link distributions deployed are the Lomax, Dagum, exponential, exponentiated exponential and the Log Logistic distributions. This technique results in new different distributions with different number of parameters, as will be discussed in the paper. For each new distribution, the author derived the PDF, CDF, quantile and hazard functions as well as the raw moments. The author expands some of the PDFs in a linear series expansion that is simple and does not require special function. This makes the computational statistics for the raw moments easier and simpler. The supplementary material associated with the paper contains figures for the PDF, CDF and the Hazard functions. Real data analysis reveals correlation between the parameters of the new distributions. The author clarified this by comparing the fitness of the new distributions to the same dataset, rate of recovery from COVID 19 in Spain, to wonder whether this correlation is data dependent or distribution dependent or mix and interaction between both the data and the distribution.

Abstract: The DiAL Health Pilot Trial is a well-designed study that measures body mass, fat mass, fat-free mass, and food intake using state-of-the-art methods. Yet its analysis converts these mass measurements into energy equivalents and uses energy as the primary variable. This commentary identifies three loci of energy derivation in the protocol: the weight-loss nomograms, the CR adherence quantification (Equation 1), and the calorie prescription itself. A concrete numerical example shows that energy-based and mass-based analyses can produce divergent results for the same participant. The energy balance equation is mathematically underdetermined with respect to growth composition: one equation, two unknowns. The mass measurements are not. The TRE arm of the trial is not affected by the same issue, because its adherence metric is time-based rather than energy-based. This makes the CR analysis an isolated case rather than a general problem with the trial. A mass-balance analysis would preserve the information that the energy conversion discards, and would produce adherence metrics directly interpretable in the units actually measured. The argument is not that the DiAL Health trial is flawed. It is that the analytical framework it inherited contains a detour that is no longer necessary given the measurements the trial already performs.

Abstract: Ensuring safety of lithium-ion batteries (LIBs) requires accounting for complex degradation mechanisms and uncertain failure conditions that can lead to thermal runaway. This study developed an integrated risk assessment methodology combining Fault Tree Analysis (FTA) with fuzzy logic to identify the main failure pathways and to estimate the likelihood of thermal runaway initiation conditions for a generic Li-ion battery system. The fault tree incorporated abuse conditions (mechanical, electrical, thermal) and cooling system faults, including accelerating mechanisms such as ageing and other stress conditions, which are often neglected in other studies. Expert judgments were translated into fuzzy sets, aggregated using the Similarity Aggregation Method, and converted into probabilistic values for top-event evaluation. The importance analysis identified BTMS failure, cycling ageing, cooling-system fouling, and high ambient temperature as the most influential modeled basic events. Comparison with previously published studies confirmed consistency in the magnitude and relevance of key failure events. The methodology provided a robust and general framework for LIB safety assessment under uncertainty data conditions, independently on specific cell chemistry or application and the framework is directly transferable to battery packs and battery management systems of electric vehicles, ships, and aircraft.

Abstract: The communication of consciousness between neurons within a single synapse is an underspecified occurrence, chemical transmission is essential for neural propagation, and neural propagation is essential within many existing measurements of consciousness. The development and plasticity of consciousness throughout an animal's lifetime requires grounded chemical distinctions of executive and non-executive processing and transmission. This paper's hypotheses propose chemical-conscious correlates between neurons, considered alongside established cognitive and environmental interaction, to specify how consciousness and its continuum are chemically produced at global scales, and developed ontogenetically during continuous environmental rationalization and neural propagation. Single-synaptic phasic transmitter transients may have evolved to allow executive function and selective consolidation within reflexive processing. These transmitter releases may be the chemical events through which subconscious stimulation allows phenomenal expression between neurons. This framing is consistent with phasic dynamics and tonic modulation of baseline synaptic stimulation, with each transmitter being differentiated by function. Phasic transients are further suggested to direct cortical spreading under tonic, epigenetic, and synaptic allowance. Because phasic releases fundamentally affect synaptic plasticity, while synapses determine the threshold and magnitude of phasic responses, this framing allows a developmental and individually variable model of consciousness. This development has potential to ground GWT/GNWT, IIT, Predictive Processing, and Affective Neuroscience, extending each framework by approaching phasic response as a mechanism of conscious variation and transmission at the scale of individual synapses. Because each hypothesis is specific and falsifiable, provided predictions and experimental directions can assist to determine the accuracy of each claim; given hypotheses are precise enough to fail and many of their correlating experiments are practical enough to proceed. If this account holds, it points toward practical improvements in depression treatments, therapeutics, memory rehabilitation, learning environment refinement strategies, and early developmental interventions.

Abstract: This study examines the relationship between underwriting concentration and financial stability indicators in Egyptian conventional property and liability insurance companies. Using a balanced panel of 17 insurers over 2020–2024 (85 firm-year observations), underwriting concentration is measured using the Herfindahl–Hirschman Index (HHI) based on written premiums across insurance lines. Financial stability is assessed using seven indicators covering profitability (ROA and ROE), liquidity (Current Ratio and Cash Ratio), and solvency/financial structure (D/E, TD/TA, and SER). Separate panel regressions are estimated using Pooled OLS, Fixed Effects, or Random Effects, as determined by specification tests. The HHI coefficient is negative for six of the seven outcomes. Under the conventional unadjusted 5% criterion, HHI is negatively associated with ROE (β = −0.0000219, p = 0.0238), while the associations with the other six indicators are not statistically significant. However, the ROE result does not remain significant under a Bonferroni adjustment for the seven HHI tests (adjusted threshold α = 0.0071). Overall, the findings do not provide robust evidence of a systematic relationship between underwriting concentration and the financial stability indicators examined, although the unadjusted ROE result provides suggestive evidence warranting further investigation.

Abstract: For \( q=2^k \) and \( \gcd(a,q)=1 \) let \( R_{a,q}(N) = \sum_{p_1+p_2=N,\ p_1\equiv a \pmod{q}} (\log p_1)(\log p_2) \) be the restricted weighted Goldbach sum, with candidate main term\( M_{a,q}(N)=2C_2/\varphi(q)\mathfrak{S}(N)N \) and exceptional set \( \mathcal{E}_{a,q}(X)=\{N\le X\text{ even: }R_{a,q}(N)=0\} \). Writing \( p_n \) for the \( n \)-th prime, \( g_n=p_{n+1}-p_n \) and \( N_n=p_n+p_{n+1} \), we prove an elementary transfer principle: if \( f \)maps a set of indices injectively (or with bounded multiplicity) into the even integers with at most linear growth, then the number of indices \( n\le X \) with \( R_{a,q}(f(n))=0 \) is at most a bounded multiple of \( \#\mathcal{E}_{a,q}(\kappa X) \). Consequences are uniform in the residue class (“ray”) modulo \( q \), and cover \( N_n=p_n+p_{n+1},\,p_n+p_{n+k} \) and the shifted primes \( p+h\ \) with \( h \) odd. A mean-value lemma of Tolev gives \( \#\mathcal{E}_{a,q}(X)\ll_{q,A}X(\log X)^{-A} \) for every \( A>0 \) when \( q=2^k \), and hence each of these sets of even integers has a restricted Goldbach representation for almost all primes in every ray, with a saving of every power of \( \log X \). We record which of these representations are trivial (a summand equal to \( p_n \) or \( p_{n+1} \)) and which are not. For a modulus q with an odd prime factor \( \ell \) we prove that \( N_n\equiv a\pmod{\ell} \) is equivalent to a condition on the pair of rays of p_n and p_n+1, that \( R_{a,q}(N_n)=O(\log N_n) \) on that set, and, by a theorem of Shiu, that this set is infinite; for q=3 we verify exactly that \( R_{a,q}(N_n)=0 \) if and only if \( N_n\equiv 1\pmod{3} \) and \( N_n-3 \) is composite. The digit-repulsion bias of Lemke Oliver and Soundararajan is documented on 5,761,451 consecutive pairs below 10⁸, together with the dependence of the mean gap on the ray of origin and a fit of the cumulative admissibility function. For q=5 the same equivalence holds with three small exceptions. We also compare the mod-3 joint frequencies of consecutive primes with the Lemke Oliver–Soundararajan prediction at three scales. Next we show that the relative second moment of \( R_{a,q}(N)-M_{a,q}(N) \) depends on \( N\bmod q \) through the symmetry \( p_1\leftrightarrow p_2 \): for \( N\equiv 2a\pmod{q} \) the variance ratio is 2.0 to 2.1 for q=8,16,32, all within about one standard error of \( 2 \). Along N_n this class effect, together with the residues of N_n modulo 3 and 5, predicts the second moment by ray of p_n to within 3% overall, and predicts the ray differences (which are significant at z=7 to 9) to within about two standard errors. Applied unchanged at q=16, with the recipe fixed in advance, it predicts the level to 2%, the spread among the eight rays to 2%, and the pattern across rays with correlation 0.96, with a residual over-prediction of 2% to 3% at all four pairs \( (q,a) \) tested, which shrinks to about 1% when the classes modulo 7 and 11 are added (exploratory). In a temporal split of the primes below 10⁸, the last ray predicts the next ray with accuracy 0.311 against 0.25 (at most 0.316 with more rays), while the rays of the past explain at most 0.19% of the variance of the normalised gap. Every statement carries an explicit status label.

Abstract:

Introduction: Continuing professional education in forensic practice requires curricula that address disciplinary expertise, interpretation, documentation, and communication while remaining responsive to changing professional needs. This scoping review aims to map how continuing professional education curricula are designed, delivered, assessed, evaluated, and quality-assured across forensic disciplines, including their relationship to ongoing professional competence. Accreditation, certification, and governance will be examined in relation to how they shape educational arrangements. Methods: The review will follow Arksey and O’Malley’s five-stage framework, incorporating Levac and colleagues’ refinements and JBI guidance. Reporting will follow PRISMA-ScR and PRISMA-S. PubMed, Embase, Scopus, CINAHL, and APA PsycINFO will be searched from inception to the final search date, with reference-list searching providing additional sources. Inclusion will be restricted to English-language publications without geographical restrictions. Eligible sources will address continuing professional education or competence maintenance among qualified or practicing forensic professionals and provide substantive information on curriculum design or educational assurance. Sources concerning only undergraduate, pre-licensure, entry-to-practice, or initial specialist education will be excluded unless a separable continuing-development component is reported. Two reviewers will independently screen titles, abstracts, and full texts using Covidence, with disagreements resolved through discussion or a third reviewer. A standardized charting form will be piloted on at least five eligible sources; one reviewer will chart each source, and a second will verify all entries. Results and Analysis: Data will be summarized descriptively and synthesized narratively across the following: learning needs assessment; competencies and objectives; curriculum content and structure; teaching and implementation; assessment and feedback; evaluation and revision; and educational quality assurance. Reported outcomes will be mapped to the four Kirkpatrick levels: reaction, learning, behavior, and results. Measurement methods and follow-up periods will be retained to distinguish immediate learning outcomes from retained competence and workplace application. The synthesis will identify shared educational elements, disciplinary variations, accessibility barriers, and reporting or evaluation gaps to inform subsequent curriculum development. Ethics and Dissemination: Ethical approval is not anticipated because the review will analyze published literature without collecting identifiable participant data. Findings will be disseminated through peer-reviewed publications, relevant professional meetings, and curriculum design summaries for educators, practitioners, professional associations, accrediting organizations, and forensic service leaders.

Abstract: Determining the precise position of a differential wheeled mobile robot requires further development of various sensor systems and localization methods. To achieve precise localization, odometry, which is based on measuring the angular velocity of its wheels, is primarily used. This article provides a brief overview of odometry methods and their calibration methods. A new method for calibrating a differential wheeled mobile robot using a trajectory measurement device is presented. The robot's trajectory measurement device is developed based on a high-precision rope displacement sensor and an absolute angular encoder. Using a trajectory measurement device during the calibration of a differential wheeled mobile robot facilitates data collection during experimental tests. The proposed method is compared with a differential wheeled mobile robot calibration method for correcting systematic errors along a linear path. The experiment was conducted on a MRTC differential wheeled mobile robot. The robot was calibrated using the developed method, and odometry error improvements were obtained, slightly better than those obtained using the well-known linear path calibration method. The method developed in this paper can be considered a simple and practical tool for calibrating various types of mobile robots.

Abstract: Apart from other vital elements of intelligent transportation and road-condition monitoring systems, vision-based pothole detection is an important one which is sensitive to the variations in illumination, especially the shadows cast by trees, buildings and vehicles. Shadows cast on the road create dark features that look like potholes, and shadows cast on top of true potholes can mask and conceal the boundaries and surface detail of the pothole, resulting in false potholes, missed potholes and less overall accuracy. In order to solve these problems, this paper puts forward a Pothole Detection Framework (SAPDF) based on the shadow awareness technology. The approach adopted is to first construct a Shadow Probability Map based on intensity, chromaticity, local contrast, texture, and edge features and then use a second stage called Shadow Pothole Discrimination to separate out illumination-induced dark regions from genuine pavement depressions. Shadows are then suppressed through an Adaptive Illumination-Guided Shadow Suppression (AIGSS) method, and the enhanced images are subsequently inputted to YOLOv11 for detection of potholes, preserving the boundaries of the potholes and the local texture. The framework is tested based on the overall detection metrics, as well as the subsets of "shadow severity". An ablation study is performed to determine the contribution of "shadow detection", "shadow discrimination", "adaptive suppression", and "structural preservation". The study will demonstrate that the explicit shadow modelling approach results in higher detection robustness under shadow-affected conditions, while preserving the detection of potholes captured under normal lighting conditions without unnecessarily changing the information, which will be the basis for reliable road-condition monitoring in complex outdoor lighting conditions.

Abstract: Explanation methods for natural language processing (NLP) come from two literatures that rarely meet: feature attribution and probing for task models, and the interpretability of large language models (LLMs), from chain-of-thought reasoning to sparse autoencoder features and circuit tracing. We place both in a unified taxonomy with four dimensions: scope (local vs. global), mechanism (how the explanation is computed), model access (model-agnostic vs. model-specific), and output form (importance scores, rules, examples, counterfactuals, rationale spans, concepts, natural language). Faithfulness and plausibility come apart, and the erasure tests built for extracted rationales do not apply to a chain of thought, which is an output rather than part of the input, so chains of thought need intervention tests of their own. The access a deployment allows settles which mechanisms are usable before scope or audience matter; our practical decision frameworks start from that question and are traced through three worked deployments. In our synthesis of LLM-era interpretability, a lineage table shows that most current techniques rebuild or extend older ideas under new constraints of scale and access, while two problems are new: keeping reasoning traces monitorable under training pressure, and testing whether models can introspect. Faithfulness metrics disagree with one another, and sparse autoencoders do not yet beat simple baselines on standardized benchmarks, so evaluations should name the tests they ran instead of reporting one score. We also survey applications across healthcare, legal and financial services, social science research, content moderation, and education, and close with six open problems.

Abstract: We propose a transport-enabled program of decoherence cartography: the inference of environment-conditioned decoherence functionals for a localized stored charged-particle qubit across distinct laboratory sites after controlling for measured technical noise. The framework is formulated generically for transportable matter and antimatter qubits, while being motivated in particular by recent progress in coherent spectroscopy of a single antiproton spin and by the physical transport of trapped antiprotons in a portable cryogenic Penning-trap. For each site and control sequence, we formulate the experiment as a hierarchical null test in which the measured coherence of the localized qubit is decomposed into a nuisance contribution predicted from monitored magnetic, vibrational, thermal, timing, and control channels, together with a residual environment-conditioned functional. In the conservative null model, transport and relocation induce only deterministic phases and standard open-system dephasing fixed by measured spectra. A residual signal that remains stable under revisits, control sequence changes, and site permutations would define an irreducible site-dependent decoherence effect and motivate targeted follow up searches for new physics, manifested in beyond the Standard Model and cosmology-motivated sectors. We formulate the operational observables, filter-function framework, identifiability conditions, dictionary examples, and benchmark sensitivities required to render both null and non-null outcomes interpretable.

Abstract: The Minimum Vertex Cover problem is NP-hard. Its classical polynomial-time approximation ratio is ~2; under \(\mathrm{P}\neq\mathrm{NP}\) no polynomial-time algorithm achieves \(\sqrt{2}-\varepsilon\) for any fixed \(\varepsilon>0\) (Khot–Minzer–Safra), and under the Unique Games Conjecture (UGC) not even \(2-\varepsilon\) (Khot–Regev). We present FindVertexCover, a linear-time ensemble algorithm that computes nine independently valid vertex covers of the input graph \(G=(V,E)\) and returns the smallest. Four of the candidates—and the redundant-vertex pruning pass applied to each—adapt the published Hvala ensemble, which proves an unconditional \(O(n+m)\) time and space bound and a worst-case ratio at most 2, where \(n=|V|\) and \(m=|E|\); Salvador therefore contains Hvala and inherits its ratio-2 guarantee. Salvador adds a Min-to-Min heuristic, a primal-dual 2-approximation, a planar forest-core reduction to a weighted Minimum Independent Dominating Set gadget solved by a Baker-style pass, and, new in this version, two further candidates: (i)~the bipartite planar oriented-incidence reduction, whose auxiliary graph we prove is always a disjoint union of even cycles and single edges, so its minimum-weight vertex cover is computed \emph{exactly} by a linear-time cycle dynamic program; and (ii)~a budgeted (1,2)-swap iterated local search that improves the best cover found and stops after \(O(n+m)\) work. The whole ensemble runs in worst-case \(O(n+m)\) time and unconditionally achieves ratio at most 2. On the 68 DIMACS clique-complement instances of the NPBench collection, Salvador never returns a larger cover than Hvala: it is strictly smaller on 48 instances and equal on the other 20, with mean ratio 1.014 against 1.025. On a reproducible suite of 1746 graphs—1718 small graphs with exact optima, all solved optimally, and 28 large adversarial graphs with up to 200,000 vertices, whose ratios are certified by matching, clique-packing and spectral lower bounds—the ratio never exceeds 1.17. We conjecture a universal ratio of at most \(7/5=\sqrt{2}-\varepsilon_0\) with \(\varepsilon_0=\sqrt{2}-7/5>0\); by the Khot–Minzer–Safra theorem a proof of this conjecture would imply \(\mathrm{P}=\mathrm{NP}\), and a proof of any constant ratio below 2 would refute UGC under \(\mathrm{P}\neq\mathrm{NP}\). These consequences are conditional: the conjecture is supported by experiments, not proved. An open-source implementation is released as the Salvador package (v0.0.9).

Abstract: Biomass cogeneration plants supplying district heating systems operate under variable thermal loads, which affect the fuel consumption attributable to electricity generation. This paper analyzes the net specific biomass consumption allocated to electricity generation in a wood biomass cogeneration plant in Suceava, Romania, with four 30 t/h steam boilers and a 29.65 MWe steam turbine. Six years of monthly operating data were used to assess the influence of the heat delivered to the district heating system and the biomass moisture content. The thermal load was the main influencing factor, since at low heat demand more steam is condensed without heat recovery. In winter months, with 1500–2500 toe of heat delivered monthly, the specific consumption ranged between 2.0 and 3.5 toe biomass/toe electricity (average ≈2.5), whereas in summer months, with 500–1500 toe delivered monthly, it ranged between 3.0 and 5.0 toe biomass/toe electricity (average ≈4.0). Biomass moisture content (30–45%) had a secondary influence. An empirical equation is proposed for the accepted limit of the annual average specific biomass consumption as a function of the average monthly thermal load, corrected by a moisture impact factor. The equation supports assessing the energy and economic efficiency of electricity generation in such plants

Abstract: The conservation of industrial heritage has shifted from mere physical preservation to adaptive reuse; accurately distinguishing between historical components, later modifications, and other architectural elements is crucial for quantitative spatial analysis and renovation assessment. However, maintaining historic character often makes elements with different semantic attributes visually similar. Key challenges include subtle local differences between similar classes, insufficient coordination between local details and global context, and class imbalance causing intra-class dispersion and inter-class overlap. Using the Tianjin First Machine Tool Fac-tory as a case study, this study constructs a 14-class foreground semantic dataset and proposes Enhanced Mask2Former for Heritage Gardens (Hg-Mask2Former++). Based on Mask2Former-Swin-B, the model introduces LGFE to enhance multi-scale local–global representations and CDCC to optimize class relationships through class-balanced sampling and supervised contrastive learning. With only a 1.49% parameter increase, Hg-Mask2Former++ achieves 73.92% mIoU and 82.67% mDice, improving the baseline by 2.12 and 1.86 percentage points. Under identical experimental settings, its mIoU exceeds that of eight comparison models, including FastVim, TPSegformer, and ViT-P-based models proposed in 2026; the IoU of Industrial Building and Modern Building improves by about 20 percentage points. Further experiments reveal class-selective and condi-tion-dependent gains, indicating that scene-specific targeted optimization is more suitable than simply increasing network complexity.

Abstract: Bailey bridges were developed as rapidly deployable modular steel bridges, but many have remained in service for decades under traffic and environmental conditions substantially different from those prevailing when they were installed. This review examines the relationship between bridge durability, historical and current truck loading, documented collapse cases, common superstructure failure mechanisms, and maintenance requirements, with evidence from South Asia, Africa, and Ethiopia. The reviewed cases indicate that, despite differences in location and service conditions, in-service Bailey bridge failures exhibit common patterns, including excessive deformation, buckling of critical members, connection failure, structural instability, and progressive superstructure collapse under vehicle loading. Overloading is frequently reported as a direct trigger, while corrosion, ageing, inadequate maintenance, flooding, and scour increase structural vulnerability. In Ethiopia, several old Bailey bridges at different locations experienced in-service collapses during 2024, 2025, and 2026, highlighting the need for systematic assessment of ageing bridges and their current loading demands. The main aim of this review is to identify and characterize common failure mechanisms of in-service Bailey bridges, raise awareness of their structural vulnerabilities, and provide a basis for preliminary investigation of old and potentially unreported Bailey bridges in Ethiopia. Construction-stage failures, such as the launching-related instability reported at the Omo Bridge, represent a separate but important failure category that also requires attention. The review therefore emphasizes condition- and load-based safety assessment through systematic inspection, corrosion and deformation monitoring, truck-load control, structural load rating, and timely strengthening, rehabilitation, or replacement.

Abstract: Type 1 diabetes mellitus (T1D) is an autoimmune disease driven by a complex interplay between genetic susceptibility, environmental factors, and dysregulated immune re-sponses, ultimately leading to the progressive destruction of insulin-producing pancre-atic β-cells. Despite substantial advances in glucose monitoring and insulin delivery technologies, current treatment strategies cannot fully reproduce the physiological regu-lation of insulin secretion and do not directly target the underlying autoimmune process. Consequently, increasing attention has been directed toward immunomodulatory ap-proaches aimed at preserving residual β-cell function and modifying disease progres-sion. Mesenchymal stromal cells (MSCs) have emerged as a promising cell-based thera-peutic approach owing to their broad immunomodulatory, anti-inflammatory, and paracrine properties. Importantly, unlike therapies directed predominantly at individual immune targets or pathways, MSCs can simultaneously modulate multiple intercon-nected mechanisms involved in T1D immunopathogenesis, affecting both innate and adaptive immune responses, inflammatory signaling, antigen presentation, and immune tolerance. This broad and multifaceted mode of action provides a strong rationale for in-vestigating MSCs as a therapeutic strategy capable of addressing the complex and highly interconnected nature of T1D autoimmunity. This review examines the major mecha-nisms underlying T1D immunopathogenesis and the multifaceted immunomodulatory effects of MSCs, followed by a critical evaluation of the available preclinical and clinical evidence on MSC-based therapy. The potential of MSCs to preserve β-cell function, mod-ulate immune responses, and improve metabolic outcomes is considered, together with current limitations and challenges for clinical translation.