Probability
Papers
Fresh from arXiv, typeset properly. Highlighted words trade on the market.
All of arXiv
Saturday 29 August
We show that the $\gamma$-LQG metric (appropriately re-normalized) for $\gamma \in (0,2)$ converges in law to the critical 2-LQG metric (appropriately re-normalized) with respect to the local uniform topology of continuous functions on $\mathbb{C} \times \mathbb{C}$. Combining with the results in arXiv:2509.15544, we obtain that the law of the $\gamma$-LQG metric (appropriately re-normalized) is continuous in $\gamma \in [0,2]$ with respect to the local uniform topology of continuous functions on $\mathbb{C} \times \mathbb{C}$, where the 0-LQG metric corresponds to the Euclidean metric on $\mathbb{C}$.
Machine Learning
Spatial Entropy based Partitioning for Spatiotemporal Graph Unlearning
Spatiotemporal graphs underpin applications such as traffic forecasting, weather forecasting, and healthcare monitoring. Privacy regulations such as the GDPR and the CCPA require the complete removal of unauthorized data from trained models, but achieving this on a spatiotemporal graph is difficult: because information propagates globally through both spatial and temporal message passing, fully erasing a node's influence forces costly full-graph retraining. ST-graph unlearning requires both exactness and efficiency. We propose IsleNet, which uses spatial-entropy-guided partitioning to create balanced, locally coherent subgraphs and reconnects them with lightweight virtual edges. Upon an unlearning request, only the affected subgraph encoder and virtual-edge layer are retrained, ensuring exact removal with low cost. Experiments on four real-world benchmarks show that IsleNet attains up to 94% of full-graph accuracy while reducing unlearning time by up to an order of magnitude. Our code is publicly available at https://github.com/wenlu-lab/STGraphUnlearning.
Read the full abstract
Systems and Control
Minimizing Grid Interconnection Capacity Requirements for AI Data Centers: A Developer-Side Planning Framework with Onsite Resources and Workload Flexibility
Securing grid interconnection capacity has become a bottleneck for AI data center projects and can take longer than constructing the facilities themselves. This mismatch can delay deployment for years, making early interconnection planning essential. This paper develops ICP-AI, an interconnection capacity planning framework from a data center developer's perspective. The framework minimizes grid import capacity under a prescribed onsite investment budget while jointly sizing photovoltaic (PV) and battery energy storage system (BESS) resources and scheduling deadline constrained workload flexibility. A secondary refinement fixes the minimum grid capacity and selects the minimum-investment PV-BESS portfolio among solutions that achieve that capacity. The framework is evaluated using monthly composite stress profiles across varying temporal assumptions, load shapes, flexible load fractions, and deferral windows. Results show that interconnection capacity reduction depends strongly on the planning environment: at a $100M budget, it is about 6% for the high load factor baseline, exceeds 10% under monthly average solar availability, and reaches 13.3% for a more diurnal load. At a $10M budget, 5% flexible load with a 1 h workload deferral window reduces BESS capacity from 15.30 to 4.87 MWh while increasing capacity reduction from 4.43% to 4.84%. To test sensitivity to temporal compression, the model is also solved over the full 8,760 h chronology, which preserves the main capacity and flexibility trends. Overall, ICP-AI quantifies the interconnection capacity and infrastructure substitution value of workload flexibility, providing an investment-interconnection frontier to support capital allocation and early project planning in constrained grid environments.
Read the full abstract
Nuclear Theory
Local-nuclear-density dependent calculation of nucleon electromagnetic form factor ratios in finite nuclei
We calculate the electromagnetic form factors of nucleons bound in finite nuclei and make predictions to the ratio between the electric ($G_E^*$) and magnetic ($G_M^*$) form factors in terms of the square of the four-momentum transfer $Q^2$. We extend our previous constant-density calculations within the covariant spectator-QMC framework by incorporating the spatial nuclear density profiles of finite nuclei and quantify the deviations from the average-density approximation used in our previous applications. The impact of the medium effects can then be observed considering the ratio between $G_E^*/G_M^*$ and the ratio in free space $G_E/G_M$, that define the proton electromagnetic double ratio associated with a given nucleus. The double ratio is expected to remove some systematic uncertainties. There is the expectation that ratios $G_E^*/G_M^*$ associated with the bound protons inside the nucleus will be measured in the near future in polarization-transfer experiments $(\vec{e} A, e'\! A' \vec{p})$. Anticipating future measurements on the subject, we make predictions for the double ratios, to be compared with the average measurements on the energy states of protons bound to nuclei. We consider the nuclei $^{12}$C, $^{16}$O and $^{40}$Ca. We conclude that the form factors calculated using nuclear density profile function $\rho(r)$ are less suppressed than in the case of the results calculated using the average nuclear density. The results indicate that the relative importance of the low-density surface region increases with $Q^2$, leading to a weaker suppression than that predicted by the average-density approximation. We also make predictions for the ratios associated with neutrons bound to nuclei.
Read the full abstract
Artificial Intelligence
LiteSearch-VL: Small Multimodal Search Agents via Trajectory Distillation and Synthetic Step-DPO
Multimodal search agents answer visual questions by interleaving image understanding, web retrieval, tool use, and evidence synthesis. Strong systems exist, but in two expensive regimes: proprietary frontier models such as GPT-5 and Gemini, or large open vision-language backbones trained with substantial agentic data and reinforcement learning. We ask a different question: when released agent trajectories are distilled into much smaller backbones under a single-node budget, what is actually transferred? We study this with LiteSearch-VL, a low-compute recipe for Qwen3-VL-2B and Qwen3-VL-4B that uses only released OpenSearch-VL trajectories, parameter-efficient LoRA adapters, and synthetic step-level preferences: DPO on GPT-5-generated hard negatives targeting five local failure modes (premature answer, wrong tool, weak query, repeated query, ignored image). Across 12,400 GPT-5-judged rollouts on SimpleVQA, FVQA, LiveVQA, and VDR-Bench-testmini, the dominant effect is behavioral rather than a uniform accuracy lift: full-trajectory supervised fine-tuning transfers the agent contract, taking the 2B model from almost never emitting a usable answer (1,237/1,240 no_answer rollouts) to 28.4% macro Pass@1, matching or slightly exceeding the off-the-shelf 4B base (25.6%). Synthetic preference learning and compact tool distillation act as refinements rather than phase transitions (best 4B configuration: 30.8% macro Pass@1). Finally, a controlled VDR step-budget ablation shows that extra search turns convert abstentions into wrong_entity errors rather than correct answers, identifying answer verification, not search depth, as the next bottleneck for small multimodal agents.
Read the full abstract
Artificial Intelligence
Plant-Inspired AI: Plants as Inspiration for Novel Problem Formulations, and Two Case Studies
Artificial Intelligence (AI) has long been inspired by studies of biological intelligence. Reinforcement learning, for instance, drew inspiration from studies involving animal learning and is now a powerful paradigm for solving many real-world problems. Recently, plant biologists have uncovered a wide range of complex behaviors in plants that enable them to flexibly adapt to variable environments. Here, we argue that such behavior can motivate new AI frameworks encompassing a range of problems overlooked by existing problem-solving frameworks such as supervised learning, tree search, and constraint satisfaction. We illustrate this idea with two examples of intelligent problem-solving in plants: (1) leaf mimicry in Boquila trifoliolata, a vine capable of altering its leaves' morphology to resemble those of multiple host trees simultaneously; and (2) coordinated root-shoot growth, wherein plants allocate resources across organ systems exploring distinct environments. While leaf mimicry is highly specific to Boquila, coordination of root-shoot growth is shared across most plants. For both examples, we capture underlying computational principles and identify problems fitting these frameworks that are currently unaddressed by AI. Finally, we outline preliminary task formulations and discuss how these formulations may be applied to non-plant problems.
Read the full abstract
Artificial Intelligence
APPSolver: Adaptive Patch Partitioning for Point-Wise Ship Flow Prediction on Unstructured Meshes
Large non-uniform point sets make direct attention-based surrogate modeling costly for ship hydrodynamics. We introduce APPSolver, a point-wise flow-prediction framework built around Adaptive Patch Partitioning (APP), a deterministic quadtree representation for fixed two-dimensional horizontal slices extracted from ship CFD simulations. APP assigns finer patches near the hull and coarser patches farther away, downsamples patch contents, and recovers predictions to the full reference point set. Under a corrected protocol that constructs natural $(t,t+1)$ pairs before splitting, reuses training-set normalization statistics, and reports three model seeds, learned tokenizers are more accurate than APP-Transformer, and a persistence baseline has lower one-step MAE on all three ShipBench hulls. The supported benefit of APP is therefore computational rather than universal predictive superiority: on a representative DTC input, APP-Transformer requires 1.815 GFLOPs and 1.309 ms per model forward, while a matched ablation shows that adaptive partitioning reduces MAE by 16.4-24.9% relative to a uniform partition augmented with learned slicing. Condition encoders provide setting-dependent gains in leave-one-hull-out evaluation, but the current absolute next-state objective does not establish accurate long-horizon dynamics. These results characterize APP as a compact spatial representation with an explicit accuracy–efficiency trade-off. Code is available at https://github.com/wenhuahuo/APPSolver .
Read the full abstract
Combinatorics
Harmonic Catalan Number Series via Half-Integer Binomial Coefficients
We develop a method for evaluating finite and infinite series involving Catalan numbers by specializing alternating binomial identities at half-integer parameters. The key ingredient is an explicit representation of the generalized binomial coefficients $\binom{m+\frac12}{k+1}$ in terms of Catalan numbers and products of odd linear factors. Combining this representation with Pascal-type identities and two identities of Bat{\i}r and Sofo, we derive several families of closed-form summation formulas involving Catalan numbers, harmonic numbers, and odd harmonic numbers. As special cases, we obtain evaluations of finite sums together with their corresponding infinite-series analogues. The approach provides a systematic mechanism for transforming identities for generalized binomial coefficients into identities for Catalan-number series.
Read the full abstract
Quantum Physics
Quantum Mechanics as a Relativistic Statistical Theory: A Mass-Shell Factorisation Approach
Spinor structure is usually introduced as part of relativistic quantum theory. We show that its underlying algebra can instead emerge without quantisation from a covariant statistical description of a massive relativistic particle subject to an explicit finite-dimensional linear factorisation of the complete massive mass-shell constraint motivated by requiring the theory retain both sheets of the massive mass shell. A finite-dimensional factorisation of the complete mass-shell constraint, linear in all four momentum components, forces a Clifford algebra. In (3+1) dimensions its minimal complex representation is four-dimensional, while each on-shell factor selects a rank-two sector; in this representation statistical completeness requires an arbitrary $2\times2$ matrix distribution within that sector before quantisation. Projecting the Weyl-ordered matrix Liouvillian gives relativistic transport while preserving all internal populations and coherences. For universal deformations on flat canonical phase space satisfying the stated covariance, matrix-neutrality, homogeneity and associativity assumptions, the resulting product is the matrix Weyl–Moyal product. Its parameter has dimensions of action and, when identified with $\hbar$, supplies the scale of physical spin-$\tfrac12$. Requiring all-order constraint preservation gives a vanishing star commutator, while preservation of the selected star-spectral sector yields the two-sided Dirac–Wigner equations. Our central conclusion is therefore that relativistic mass-shell factorisation supplies spinor structure, while completion of the constraint-preservation hierarchy within the stated deformation class leads to the Weyl–Moyal phase-space formulation of quantum mechanics.
Read the full abstract
Artificial Intelligence
Cross-Relational Preference Learning for Better LLM Instruction Following
Large Language Models (LLMs) still exhibit limited capability in following complex instructions. While existing approaches often rely on preference learning to enhance this ability, they typically overlook the relationships between the permissible response spaces of different instructions, which restricts a model to align with subtle and diverse constraint variations. To address this, we propose Cross-Relational Preference Learning (CRPL), a novel framework for constructing preference data that explicitly models inter-instruction relationships through two key techniques: Cross-Relationship Perturbation and Cross-Region Pair Sampling. This enables the generation of more diverse preference data that captures a wide spectrum of constraint variations. Additionally, we introduce an atomic constraint-based verification mechanism to rigorously assess response satisfaction, ensuring high-quality preference pair construction. Extensive experiments across multiple preference learning methods (e.g., DPO, KTO), LLM backbones and four instruction-following benchmarks demonstrate that our approach achieves substantial improvements over prior baselines and exhibits strong generalization.
Read the full abstract
Geometric Topology
Rigidity of the period map up to finite covers
We first give a complete classification of bi-affine representations of mapping class groups of surfaces with finitely many boundary components or punctures. We also show that every linear representation of the mapping class group of a genus-$g$ surface with two boundary components of dimension at most $3g-3$ is bi-affine. We then classify low-dimensional symplectic representations of the mapping class group associated to triple covers. Let $[\beta]\in H_1(S_g;\mathbb{Z}/3\mathbb{Z})^*$, and let $\widetilde{S}\to S_g$ be the corresponding triple cover with deck transformation $\sigma$. For $h\le g$, every non-abelian homomorphism from either $\mathrm{Mod}(S_g,[\beta])$, the stabilizer of $[\beta]$ in $\mathrm{Mod}(S_g)$, or $\mathrm{Mod}(\widetilde{S},\sigma)$, the centralizer of $\sigma$ in $\mathrm{Mod}(\widetilde{S})$, to $\mathrm{Sp}_{2h}(\mathbb{Z})$ is, up to conjugation, the standard symplectic representation on $H_1(S_g;\mathbb{Z})$. As an application, we obtain a rigidity theorem for holomorphic maps from the moduli space $R_g^{(3)}$ of genus-$g$ curves equipped with a $3$-sheeted (unbranched) normal covering to the moduli space $\mathcal{A}_h$ of $h$-dimensional principally polarized abelian varieties. We prove that, for $g\ge 6$ and $h\le g$, the unique nonconstant holomorphic map from $R_g^{(3)}$, equipped with either of its two natural complex-orbifold structures, to $\mathcal{A}_h$ is the period map sending a cover $Y\to X$ to the Jacobian of the base curve $X$.
Read the full abstract
Quantum Physics
Quantum Natural Gradient on Quotient Spaces
A parametrized quantum circuit reports its state geometry through a quantum Fisher information matrix (QFIM), often singular. A small Fisher value can reflect exact state-preserving redundancy, compression by the circuit chart, or weak intrinsic distinguishability, and these mechanisms call for different numerical treatments. We show that the circuit metric factors as $F=B^{*}MB$, where $B$ is the state-level circuit-to-orbit differential and $M$ is the intrinsic Fisher operator on the reachable orbit. The factorization identifies the exact kernel as $\ker B$, separates coordinate transfer from intrinsic geometry, and yields the condition for a circuit to realize an orbit-level quantum natural-gradient (QNG) direction. When the prescribed redundancy exhausts the Fisher kernel, the Moore–Penrose update is the minimum-norm horizontal lift of the quotient Riemannian gradient. At critical points with a locally diffeomorphic quotient-to-orbit map, chart singular values cancel from the linearized QNG operator while intrinsic anisotropy remains; in the trace-orthonormal full-control generator frame, excitation-gap anisotropy gives $\kappa_{\mathrm{QNG}}=\kappa_{\mathrm{Eucl}}$. Representation theory makes $M$ explicit on highest-weight, Slater, and fermionic-Gaussian orbits, and cominuscule fidelity flow becomes integrable, with conserved principal-defect ratios, cubic Lie-retracted convergence at $\eta=2$, and stability boundary $\eta=4$. Finite data impose a second boundary: an estimated QFIM and its confidence radius alone cannot distinguish an exact zero from a small physical mode, so the estimated spectrum alone cannot license hard projection. Under depolarization, inverse-Fisher scaling amplifies mean updates and fluctuations together and cannot restore update signal-to-noise. A redundant Slater/Givens circuit confirms exact transfer identities and illustrates finite-shot tradeoffs.
Read the full abstract
Machine Learning
Information-Based Calibration of Uncertainty Quantification in Product-of-Experts Gaussian Process Models
Gaussian process (GP) regression with a single global GP (GP-glo) incurs cubic computational cost, limiting scalability to large datasets. Product-of-experts GP models (GP-pro), which combine local GP models to capture global correlations, alleviate this computational burden. However, training local experts on disjoint data subsets can lead to overestimated posterior variances. We propose GP-pro-c, a product-of-experts GP model that calibrates these variances using an information-based method. The method exploits the monotonicity and submodularity of information gain in GPs to define a calibration ratio that reduces the posterior variance of individual local GP models. We evaluate GP-pro-c using negative log-likelihood (NLL), root mean squared error (RMSE), and expected normalised calibration error (ENCE). Experiments on four synthetic functions and six regression datasets show that GP-pro-c achieves average reductions of 2.3% in NLL and 12.0% in ENCE compared with the uncalibrated GP-pro model. The proposed method mitigates posterior variance overestimation while maintaining predictive accuracy and reducing computational complexity. GP-pro-c provides a promising approach for uncertainty estimation in scalable GP models and may serve as a useful surrogate model for Bayesian optimisation with high-dimensional and large-scale data.
Read the full abstract
Artificial Intelligence
Extending TotalSegmentator: Predicting Patient and Acquisition Characteristics from CT and MR Images
Background: Patient details and acquisition metadata are important for clinical decisions, image quality control, and automated research pipelines, but may be missing or unreliable in imaging archives. Purpose: To develop and evaluate a fast open-source model that predicts patient and acquisition characteristics directly from CT and MR images. Materials and Methods: Separate 3D ResNet-10 ensembles for CT and MR were trained on 57,291 and 43,200 clinical examinations acquired from 2011 to 2025. Both predicted weight, height, age, sex, contrast presence, vertebral coverage, and image noise. The CT model additionally predicted scanner manufacturer, tube voltage, tube current, convolution kernel, and post-injection time; the MR model predicted sequence class. Performance was evaluated on internal CT (n=501) and MR (n=636) test sets and an external CT dataset (n=54). Results: Internal CT MAEs were 3.90 kg, 3.68 cm, and 4.42 years for weight, height, and age, with sex F1=0.990; corresponding MR results were 4.34 kg, 4.62 cm, 7.13 years, and F1=0.970. The CNN outperformed a segmentation-derived XGBoost baseline for all four core targets in both modalities (adjusted P<=.042). F1 scores were 0.963 for CT contrast, 0.953 for MR sequence, and 0.823 for MR contrast. External CT MAEs were 4.45 kg, 4.05 cm, and 5.17 years, with sex F1=0.971. CPU inference required 20 seconds for CT and 12 seconds for MR. Conclusion: One 3D multitask model per modality can rapidly recover patient and acquisition characteristics from heterogeneous CT and MR examinations. Models are available in TotalSegmentator: https://github.com/wasserth/TotalSegmentator
Read the full abstract
Robotics
A Cognitive Architecture for Shared Autonomy in AUV Operations
Operators remain essential to Remotely Operated Vehicle (ROV) operation, yet often suffer from low situational awareness and high workload, both of which negatively affect safety. This paper presents a cognitive architecture consisting of an ontology and multiple Large Language Models (LLMs) to assist the operator at all stages of the mission. Each LLM is grounded with domain-specific information from the ontology and given a simple role to create a system that can support the operator at all stages of an operation. We are aiming to prove that using the two together will allow decisions to be grounded in the relevant domain knowledge, but also benefit from the reasoning capabilities of the LLM. Our framework determines if a mission is possible for a given Unmanned Underwater Vehicle (UUV), performs mission planning, and executes a given mission in simulation. The operator can be involved in planning and execution, ensuring the resulting plan is valid and that the vehicle behaves safely during execution. We compare different LLMs, Llama3, GPT-OSS, and Qwen2.5, to determine which are best suited to the different roles within our framework. We find that GPT-OSS performs best for feasibility assessment, planning, and execution, while Qwen2.5 is best suited to identifying mission types from natural language input.
Read the full abstract
Computer Vision
GSPotential: Camera Potential Field for Sparse-View 3D Gaussian Splatting
3D Gaussian Splatting has achieved remarkable success in photorealistic rendering, yet it suffers from severe overfitting and geometric artifacts in sparse-view scenarios due to the inherent deficiency of photometric supervision. Recent advances have attempted to regularize optimization by incorporating external priors, such as depth, point clouds, or diffusion models. However, these methods typically overlook the non-uniform distribution of supervision across the viewing space, resulting in limited specificity in prior use and primitive control. In this paper, we propose GSPotential, a framework that quantifies view-space supervision imbalance using a Camera Potential Field. Our key insight is to identify supervision valleys where photometric constraints are most deficient, and use the potential field to guide reconstruction from two complementary aspects. First, we devise a probabilistic spherical sampling strategy that places informative virtual cameras in low-potential regions. Point-cloud renderings from these views then provide targeted geometric guidance. Second, the same field provides a directional coverage cue for conservative Gaussian updates in weakly covered spatial sectors. Extensive experiments demonstrate that GSPotential achieves high reconstruction fidelity while maintaining competitive training efficiency.
Read the full abstract
Artificial Intelligence
BIRD-History: A Benchmark for History-Driven Text-to-SQL with Fine-Grained Knowledge Annotations
While recent Large Language Model (LLM)-based text-to-SQL systems achieve impressive performance on standard benchmarks, they struggle when user queries implicitly rely on domain-specific knowledge, such as business logic, data conventions, and analytical practices, that is neither captured by the schema nor explicitly stated in the natural language question. Historical SQL query logs offer a valuable source of such knowledge, yet existing benchmarks do not adequately support evaluation of history-driven approaches. To address this gap, we introduce BIRD-History, a benchmark consisting of 1,393 tasks across 11 databases, designed to evaluate text-to-SQL systems' ability to ground underspecified natural language questions using historical SQL scripts. Each task is annotated with ground-truth labels specifying which historical queries contain relevant knowledge and which SQL clauses encode it, enabling systematic evaluation of both retrieval effectiveness and knowledge utilization. Alongside the benchmark, we propose a plug-in retriever that extracts five types of external knowledge from historical SQL scripts, then retrieves and reranks relevant fragments for query generation. The retriever integrates seamlessly into existing few-shot text-to-SQL pipelines without requiring prompt modifications. Experiments demonstrate consistent improvements across four text-to-SQL systems, highlighting the value of leveraging historical query logs for handling underspecified queries. Dataset and code are open-sourced on https://github.com/zjuidg/BIRD-History.
Read the full abstract
Chemical Physics
Transition states and dynamical bottlenecks in the Li-NC <-> Li-CN isomerizing reaction with the Lagrangian betweenness
We study the vibrational dynamics that is responsible for Li-NC <-> Li-CN isomerization reaction using the Lagrangian betweenness, a tool recently developed to identify dynamical bottlenecks in the context of oceanographic transport. Following the spirit of the betweenness-centrality metric, which efficiently identifies the nodes that act as bottlenecks in complex networks, the Lagrangian betweenness is similarly capable to identify bottlenecks in dynamical systems. On the one hand, in this paper we show the ability of the Lagrangian betweenness to identify transition states, which are short-lived states that are formed at the top of the energetic barrier that the system must surpass in order to transition the reaction from the (initial) reactants to the (final) products. On the other hand, we also demonstrate that the Lagrangian betweenness is also able to identify other dynamical bottlenecks, and phase-space invariant structures such as manifolds, chains of islands and invariant tori, which are also responsible for the intricate vibrational dynamics of the system.
Read the full abstract
Soft Condensed Matter
"Ultra-weak" First-Order Phase Transition in Biaxial Liquid Crystals
We report high-resolution Monte Carlo evidence characterizing the hierarchy of transition strengths in biaxial nematogens. By employing multiple histogram reweighting on large lattices, we demonstrate that while the isotropic-to-uniaxial ($I \to N_U$) transition is symmetry-enforced first-order, the uniaxial-to-biaxial ($N_U \to N_B$) transition in the region between the tricritical point and the triple point ($\lambda = 0.26$) is proximity-driven and more than an order of magnitude smaller for large system sizes. Such a class of ultra-weak first-order transitions could be a new pathway for future investigations.
Read the full abstract
Fluid Dynamics
Droplet impact of emulsions on inclined microstructured surfaces
Droplet impacts rarely occur on horizontal smooth surfaces, yet the combined influence of surface inclination, microstructure, and fluid composition remains poorly understood for emulsions. Here, we experimentally investigate the impact of 20% (v/v) silicone oil-in-water emulsion droplets on smooth glass and micropillar arrays with two interpillar spacings (50 and 100 $\mu$m) inclined at 55$^\circ$. The role of surface wettability is further examined using a superhydrophobic coating. High-speed imaging is employed to characterise the impact dynamics over a wide range of Weber numbers ($18 \leq We \leq 565$), while optical microscopy is used to quantify the residual wetted area after impact. Five impact regimes are identified: deposition, sliding, partial rebound, complete rebound, and splashing. Surface inclination introduces asymmetry in spreading and recoil, whereas microstructured substrates reduce adhesion and promote rebound by decreasing the liquid-solid contact area. In contrast to water, the emulsion exhibit enhanced deposition on smooth surfaces, increased sliding on uncoated microstructures, and earlier ligament formation and secondary breakup at high Weber numbers. The dispersed oil phase locally modifies wetting and lubrication, producing liquid bridges, heterogeneous recoil, and larger residual wetted areas on smooth substrates. On microstructured surfaces, however, capillary confinement within the texture limits the projected wetted area despite increased liquid retention inside the microstructure. These results demonstrate that the impact behaviour of emulsions cannot be predicted solely from conventional dimensionless groups, as local phase redistribution fundamentally alters energy dissipation, wetting, and fragmentation. The findings provide new insight into spray deposition on inclined textured surfaces relevant to agricultural spraying and coating technologies.
Read the full abstract
Applied Physics
Advancing Lead-Free FASnI3 Perovskite Solar Cells via Cu-Doped NiOx HTL Engineering: An Optoelectronic Analysis
Efficient lead-free perovskite solar cells remain limited by inefficient charge transport and significant recombination losses. In this work, copper (Cu) doped nickel oxide (NiOx) is employed as the hole transport layer (HTL) in a formamidinium tin iodide (FASnI3) perovskite solar cell to improve hole extraction and suppress carrier losses. The optoelectronic response of the proposed device architecture is evaluated through a coupled numerical framework combining finite-difference time-domain (FDTD) optical simulations with finite element method (FEM)-based electrical modeling. Electrical analysis solves the Poisson equation together with the drift-diffusion and carrier continuity equations, while accounting for radiative, non-adiative, and interfacial recombination mechanisms. Systematic optimization of the thicknesses and doping concentrations of the constituent layers was performed to establish a high-performance device configuration. Following optimization of layer dimensions and carrier densities, the device achieved a power conversion efficiency (PCE) of 24.71%. Integrating an anti-reflection coating (ARC) effectively suppressed reflection losses, yielding a 5.47% relative enhancement in PCE. Consequently, the optimized device achieved a peak PCE of 26.06%, with a short-circuit current density (Jsc) of 28.98 mA/cm2, an open-circuit voltage (Voc) of 1.051 V, and a fill factor (FF) of 85.64%. These findings demonstrate that Cu-doped NiOx functions as a highly promising HTL to boost charge collection and maximize overall efficiency in lead-free FASnI3 solar cells.
Read the full abstract
Materials Science
Strain-Driven Electronic and Catalytic Modulation of g-C3N4/GeS van der Waals heterostructure for Photocatalytic Water Splitting
Photocatalytic water splitting offers a viable pathway for sustainable hydrogen production. In this study, first-principles density functional theory calculations were performed to explore the strain-dependent photocatalytic behaviour of a two-dimensional g-C3N4/GeS heterostructure. The heterostructure shows type-II band alignment with an indirect band gap of 2.17 eV, smaller than those of the individual g-C3N4 (2.81 eV) and GeS (3.31 eV) monolayers. Biaxial tensile strain up to 3% effectively modulates the band gap and band edge positions, allowing suitable alignment with water redox potentials. The calculated Gibbs free energy for the hydrogen evolution reaction ({\Delta}GHER) is 0.2 eV for the pristine heterostructure and approaches near-thermoneutral values (-/+ 0.1 eV) under +1% and +2% strain. Meanwhile, the OER overpotential decreases from 2.17 V to 0.97 V with increasing strain. AIMD simulations and optical absorption in the visible region confirm the thermodynamic stability and promising photocatalytic potential of the heterostructure for hydrogen generation.
Read the full abstract
Mathematical Physics
Sharp Logarithmic Quantum Dynamics for Quasiperiodic Schrödinger Operators
Dynamical localization requires all position moments of a quantum wavepacket to remain bounded in time, but for quasiperiodic Schrödinger operators such bounds are generally not uniform in phase. In the positive Lyapunov exponent regime, the best known phase-uniform estimates instead grow on a logarithmic scale. We prove that both the logarithmic scale and the dependence on the moment order are sharp for a class of one-frequency quasiperiodic Schrödinger operators with even potentials. Our main ingredient is a reflective version of semi-uniformly localized eigenfunctions, adapted to the two localization centers forced by a completely resonant phase, from which we obtain matching logarithmic lower bounds along sequences of times.
Read the full abstract
Materials Science
Analogue Phase Change Computational Memory with High Precision Reads and Energy Efficient Writes
Resistive memory technologies offer a compelling advantage for in-memory computing. However, realizing a device architecture that simultaneously achieves high computational precision, efficiency, and density has remained elusive due to inherent trade-offs among these performance metrics. Here, we introduce a com- pact phase-change memory device architecture that combines an ultra-confined active switching volume for enhanced electro-thermal efficiency with a non-insulating thin film that suppresses temporal conductance fluctuations. We analytically model and back-end integrate these devices into crossbar arrays. Even with conventional, undoped phase-change materials, the device architecture enables a computational precision approaching 6 bits, low conductance values below 50 {\mu}S with an adequate conductance window, and a viable pathway toward sub-100 {\mu}A programming currents under nominal operating voltages.
Read the full abstract
Number Theory
Vanishing of $\mathrm{Sha}(A/K)[\mathfrak{P}^\infty]$ and its consequences for the anticyclotomic Iwasawa theory of $\mathrm{GL}_2$-abelian varieties
In this article we generalise a classical Kolyvagin's result on the vanishing of the $p$-part of the Shafarevich–Tate group of an elliptic curve (defined over $\mathbb{Q}$) over an imaginary quadratic field $K$ to modular abelian varieties of $\mathrm{GL}_2$-type (defined over a totally real number field $F$) and a CM field $K/F$. Combining this result with the abstract Iwasawa-theoretical argument of [MN19], we show that a similar vanishing holds over the layers of a suitably defined anticyclotomic multi-$\mathbb{Z}_p$-extension of $K$.
Read the full abstract
Computer Vision
FISICA: A Deployed Service for Plantar-Pressure and Posture Assessment with Ontology-Grounded Recommendation
FISICA is a body-assessment and recommendation service running in production. One standing session with two photographs returns foot-loading measures, posture coordinates, a driven 3D avatar, a visual report, and ranked shoe and exercise candidates. Measurement comes from a purpose-built scale carrying 634 force-sensitive elements on a 1 cm grid and four load cells, and a rule-based evaluator controls every recommendation while a language model only explains the stored result. The method contribution is the avatar. Instead of mapping a measured angle onto a rig through a tuned gain, we measure the avatar with the same function used on the subject and solve until the two agree, on a sampling-invariant spinal metric that separated a normal from a kyphotic record by 7.2 degrees against 0.9 degrees for a single-joint formulation. In production, general APIs respond at a 0.023 s median, plantar-pressure analysis at 0.45 s, and recommendation at 2.16 s to 2.26 s with the rule-based portion under one second in every trial. The served keypoint graph reaches 0.960 PCK@0.2 on public data, and the catalog holds 699 shoes with 10,500 typed facts. An approved study supplies the radiographic reference for the validation still ahead.
Read the full abstract
Computer Vision
GenFirst: Generation Before Reconstruction for Stable End-to-End Latent Generative Modeling
Latent generative models typically follow a two-stage pipeline, training a variational autoencoder for reconstruction and then a generative model on the frozen latent space. Since reconstruction-optimized latents are not necessarily generation-friendly, jointly training both models is an appealing alternative. However, direct end-to-end training remains challenging, as it is prone to latent collapse and faces a generation-reconstruction conflict. We revisit this problem by analyzing how different objectives shape the latent space and identify two key insights. First, the entropy term in the Kullback-Leibler divergence objective is essential for preventing collapse: reconstruction and prior fitting tend to shrink the posterior, while entropy preserves non-degenerate latent uncertainty. Second, reconstruction and generation exhibit asymmetric learning dynamics: reconstruction is fast and strongly supervised, whereas generation is slower and harder to optimize. Based on these insights, we achieve the first direct end-to-end training without latent collapse and propose GenFirst, a simple generation-before-reconstruction strategy. The generative objective first shapes the latent space under weak reconstruction pressure, after which reconstruction is progressively strengthened to recover visual details. We validate GenFirst with continuous autoregressive priors with exact likelihoods and SiT priors with implicit likelihoods. With our end-to-end objective and GenFirst, SiT achieves a gFID of 0.97 with CFG and 1.45 without CFG on ImageNet-256, while MMDiT reaches a GenEval score of 0.90 on text-to-image generation. Beyond image generation, we extend the framework to shared visual latents for generation and representation learning, and to continuous unified text-image generation. These results demonstrate the generality of stable end-to-end latent learning across generative priors and modalities.
Read the full abstract
Solar and Stellar Astrophysics
On the Effect of Large Scale Structures and Turbulence on Solar Eruptive Events: The Cross-Scale Challenge for Space Weather
Coronal Mass Ejections (CMEs) are among the most powerful drivers of space weather, yet their prediction remains elusive. A fundamental obstacle is the problem's multiscale nature: large-scale magnetohydrodynamic models do not resolve the turbulent fluctuations that govern particle transport and magnetic connectivity. We present a new model that combines a MICroscopic diffusion approach of turbulence with MACroscopic EUHFORIA simulations (MICMAC) to address the cross-scale challenge of space weather. By incorporating turbulence properties into such small-scale Monte Carlo simulations, we describe an extreme CME event, tracing both magnetic field lines and 100 MeV protons from the CME-driven shock to 1 AU. We find that turbulence dramatically broadens and distorts the magnetic connection between the CME nose and Earth, producing footpoint distributions that are highly non-Gaussian, anisotropic, and patchy. MICMAC suggests that the distribution's enstrophy (non-Gaussianity) grows as the CME approaches Earth, signaling anomalous diffusive behavior. A simple shear-layer toy model reproduces the observed in-plane anisotropy, suggesting that local current sheet geometry imprints a persistent memory on the turbulent field. Our results demonstrate that cross-scale coupling between the CME's large-scale structure and ambient turbulence must be accounted for in space weather models. We discuss implications for SEP forecasting and interpreting multi-spacecraft observations.
Read the full abstract
Quantum Physics
Zero-Energy Problems for Supersymmetric Hamiltonians on a Chain Are $\mathrm{QMA}_1$-Complete
We determine the computational complexity of deciding whether a supersymmetric quantum system on a one-dimensional chain has a zero-energy ground state. We prove $\mathrm{QMA}_1$-completeness for two input models, formulated in terms of Hermitian and nilpotent supercharges, respectively. In the Hermitian model, both the supercharge terms and the Hamiltonian have bounded range on the chain; in the nilpotent model, this geometric restriction applies only to the Hamiltonian. The hardness reduction from quantum satisfiability on a chain preserves exact zero energy and gives an inverse-polynomial ground-energy lower bound for negative instances. A common verification method establishes membership in $\mathrm{QMA}_1$ while allowing cancellations between supercharge terms.
Read the full abstract
Analysis of PDEs
Global recovery of Lorentzian principal geometry from the hyperbolic Dirichlet-to-Neumann map
We prove that the full Dirichlet-to-Neumann map on a finite time interval globally determines the Lorentzian metric \(\mathbf G_{g,c}=-c(x,t)^2\mathrm{d}t^2+g(x)\) encoded by the principal symbol of the wave operator. The result holds in spatial dimensions \(n\geq3\) for multiplicatively separable wave speeds \(c(x,t)=a(x)b(t)\), provided the accumulated effective time during the experiment exceeds the maximal travel time from the boundary to the interior and back. The data determine both the unknown Riemannian metric \(g(x)\) and the wave speed \(c(x,t)\) throughout the observation cylinder, up to a spatial change of coordinates that fixes the boundary and leaves the measured time unchanged. In particular, the temporal factor \(b\) is not prescribed but is recovered from the same measurements. We also construct several examples demonstrating that the dimensional and visibility conditions for uniqueness are sharp.