Peer Revenue Equity in science

Papers

Fresh from arXiv, typeset properly. Highlighted words trade on the market.

All of arXiv

Friday 28 August

Systems and Control

Scalable Voltage-Stability Dataset Generation Via Boundary-Proximity Indicators Clustering

This paper proposes a scalable framework for voltage-stability dataset generation. Voltage-stability-constrained planning increasingly relies on machine-learning surrogates but training them requires large datasets labelled by continuation power flow (CPF) results, which is computationally costly. To address this, this paper proposes a framework that uses hierarchical clustering on boundary-proximity indicators to reduce the number of required CPF evaluations. The proposed approach combines (i) uniform sampling of feasible operating space using Hit-and-Run Markov Chain Monte Carlo, (ii) structured stress directions via maximin Latin hypercube sampling (LHS), (iii) sensitivity-guided perturbations to target weak buses, and (iv) clustering-based representative CPF labelling that reconstructs the voltage stability margins of unlabelled operating points from representative cluster medoids. Results on the IEEE 39-bus system show that the proposed framework significantly reduces CPF evaluations by 95.45% while preserving high accuracy and boundary fidelity for both regression and classification tasks. The reduced surrogates remain structurally consistent with their full-CPF dataset counterparts, demonstrating the suitability and scalability of the proposed approach for operation and planning optimization.
Read the full abstract

Nuclear Experiment

The role of strangeness and baryon enhancement in heavy-quark hadronisation from pp to Pb$-$Pb collisions with ALICE

In this contribution, a combined study of charm-meson and charm-baryon production with ALICE is presented. The $\Lambda_{\mathrm{c}}^{+}/\mathrm{D^{0}}$ ratio in pp collisions at $\sqrt{s} = 13.6$ TeV and the first measurement of the nuclear modification factor of $\Xi_{\mathrm{c}}^{0}$ baryons in Pb$-$Pb collisions at $\sqrt{s_\mathrm{NN}} = 5.02$ TeV are reported. Furthermore, measurements of the $\mathrm{D_s^+/D^+}$ production ratio from low-multiplicity pp collisions at $\sqrt{s} = 13.6$ TeV to Pb$-$Pb collisions at $\sqrt{s_\mathrm{NN}} = 5.36$ TeV are presented. Lastly, the first measurement of the $p_\mathrm{T}$-differential $\mathrm{D_{s1}(2536)^+/D_s^+}$ yield ratio is shown.
Read the full abstract

Systems and Control

Hierarchical Agglomerative Clustering for Efficient Annual Voltage Security Assessment in Very-High RES Penetrated Power Systems

Voltage security assessment in power systems with high renewable energy source (RES) penetration requires analyzing many operating conditions to capture variability and uncertainty, but simulating a full year of operating points is computationally costly - motivating the selection of representative operating points (ROPs). Most existing methods cluster demand and generation profiles, but similarity in these profiles does not guarantee similarity in voltage behavior, since reactive power limits, voltage-control actions, and nonlinear network interactions shape voltage response in ways that cannot be inferred from power profile patterns. This paper proposes an unsupervised learning framework that selects ROPs based on the system's actual voltage response: each operating point is represented by system-wide voltage-risk indices from AC power-flow solutions, Principal Component Analysis reduces dimensionality, and Hierarchical Agglomerative Clustering with Ward linkage identifies representative voltage regimes. A comprehensive set of evaluation criteria then measures how well the selected ROPs reproduce the full year's voltage-security characteristics under normal and contingency conditions. On the IEEE Voltage Test System under very high RES penetration, the framework reduces the annual operating point set by 99.66 percent while reproducing full-year voltage behavior with 98.3 percent reconstruction accuracy in steady state and 93.4 percent in post-contingency response, outperforming existing injection-space clustering and heuristic sampling.
Read the full abstract

Artificial Intelligence

Memristive-Friendly Hadamard Reservoir Computing: Structured, Multiplier-Free Recurrences at Scale

Reservoir Computing (RC) designs Recurrent Neural Networks around a fixed, i.e., untrained, recurrent layer, and is a natural candidate for neuromorphic hardware. Memristive-friendly reservoirs derive the neuron dynamics from memristive-device kinetics, but still rely on dense recurrent matrices, which are expensive to realize physically. In this paper, we replace the dense matrix with a structured orthogonal operator, built from sign diagonals, a permutation, and a fast Walsh-Hadamard transform. The operator is multiplier-free, requires $O(N)$ parameters and $O(N\log N)$ operations per step, and is never materialized as a matrix. We instantiate it in a standard and in a memristive-friendly Echo State Network, with one binary input connection per unit. Our mathematical analysis shows that exact orthogonality yields an echo state condition that is tight in the recurrent scaling, and a noise response that is predictable at design time. Moreover, the operator mixes the whole state in a single application. Experiments on twenty classification and seven regression benchmarks, at reservoir sizes up to $N = 8192$, show that the structured models match dense orthogonal reservoirs, and achieve better mean performance than the cycle reservoir by a margin that widens with size. Furthermore, we time the recurrent step on three hardware platforms, where it is up to $50\times$ faster than a dense product and $10^4\times$ smaller in memory. Finally, we ablate the operator and measure the response to noise, quantization, device mismatch and discrete faults.
Read the full abstract

All authors Andrea Ceni, Gianluca Milano, Carlo Ricciardi, Claudio Gallicchio

Fluid Dynamics

LES of iron-powder combustion in a jet-in-hot-coflow burner - Insights on flame structure and ignition characteristics

Micron-sized iron powders are a rapidly advancing novel energy storage technology. In order to improve the design of real-world iron-powder combustors, adequate understanding of the ignition behavior in such settings is necessary. In this work, a jet-in-hot-coflow (JHC) burner designed by Hameete et al. (2024) to test ignition of iron particles in lab-scale turbulent flames is modeled numerically using LES and Lagrangian point-particles. The JHC burner is simulated in two different modes–an open flame and an enclosed flame–similar to the experimental reference. Two iron-oxidation-rate models–the first-order model and oxide-layer model–are used to examine the effect on capturing the ignition behavior. For the radiative heat transfer between the two phases, a simplified Stefan-Boltzmann approximation model and the P1 model are considered, similar to Ramaekers et al. (2025). Analysis of flame structure indicates ignition in the circumference of the jet, aided by the break-up of the coflow. At higher $T_\mathrm{coflow}$, ignition onset and oxidation completion is earlier prior to jet break-up. Minimum coflow temperature for particle ignition with the oxide layer model is $1125\mathrm{K}$ with complete oxidation at $1250\mathrm{K}$, and for the first-order model at $800\mathrm{K}$ and $900\mathrm{K}$, respectively. Both of these results do not match the experimental results of Hameete. Oxidation degree is predictably higher for enclosed flames. For the chosen particle distribution, the P1 model exhibits higher radiative heat loss and results in a slightly lower oxidation degree. Analysis on particle ensembles with partial oxidation shows that the overall oxidation degree at a sufficient height above the nozzle reflects particle ignition probability. Further analysis in regards to particle size shows ignition failure is more prevalent in larger particles.
Read the full abstract

Computation and Language

A Probabilistic Interpretation of KV Cache Eviction

The premise and promise of KV (cache) eviction is simple: higher throughput can be achieved by evicting some entries from the KV cache, at a negligible cost to quality. This holds empirically for many existing methods, though most rely on creative heuristics for selecting which entries to drop. Despite recent advances, the problem of KV eviction has remained informal in the literature. This paper aims to properly formalize this problem through the lens of probabilistic reasoning and reveal what can be learned from this perspective. Concretely, we (1) formalize the problem of KV eviction and, unfortunately, prove that it is computationally hard, (2) show that by framing it probabilistically, KV eviction reduces to the problem of expectation estimation, which can be approximated through sampling, (3) show that through this probabilistic interpretation, correcting for evicted entries during decoding—a previously ignored problem—becomes feasible, and (4) reveal that existing methods in the literature are zero-variance biased estimators that can be easily adapted in order to enable decode time correction. In practice, we show that this probabilistic version of KV eviction coupled with decode time correction is more robust to different tasks compared to existing eviction methods and achieves competitive performance at the same compression budget.
Read the full abstract

All authors Renato Geh, Alex Chen, Daniel Israel, Aditya Grover, Guy Van den Broeck

Analysis of PDEs

Characteristic localization, sharp lifespan asymptotics and global dynamics for a one-dimensional derivative wave equation

We study the Cauchy problem $$ v_{tt}-v_{xx}=\abs{v_t+v_x}^{m}\abs{v_t}^{n}, \qquad v(x,0)=\eta\varphi(x),\quad v_t(x,0)=\eta\psi(x), $$ on $\R$, with compactly supported profiles $\varphi\in C_0^2(\R)$, $\psi\in C_0^1(\R)$, exponents $m,n>1$, and amplitude $\eta>0$. We show that the sign of $P_0=\psi+\varphi'$ decides the behaviour of small solutions, whatever the sign of $\psi-\varphi'$. If $P_0$ is positive at some point, the lifespan $T(\eta)$ obeys explicit two-sided bounds of order $\eta^{-(m+n-1)}$ for every $\eta>0$, and $$ \lim_{\eta\to0^+}\eta^{m+n-1}T(\eta)=\frac{2^{n}}{(m+n-1)\,(\max_{\R}P_0)^{m+n-1}} . $$ If $P_0\le0$, the solution is global for every amplitude below an explicit threshold. If moreover $P_0\not\equiv0$, the component $v_t+v_x$ decays at the universal rate $\bigl(2^{n}/((m+n-1)t)\bigr)^{1/(m+n-1)}$, and the gradient of the solution converges uniformly to that of a free wave travelling to the right; if $P_0\equiv0$, the solution is itself such a travelling wave. The analysis rests on a localization property: the zero set of $v_t+v_x$ is invariant along its own characteristics, so that the nonlinear source stays in a slab of fixed width moving with speed one, and each characteristic of the other family is forced only during a bounded time. This yields the global existence, the long-time behaviour and the exact value of the limit. The results extend to the endpoint exponents $m,n\ge1$ and to sources $f(v_t+v_x)g(v_t)$, for which $T(\eta)$ is asymptotic to an Osgood-type integral.
Read the full abstract

2608.28292 · 28 Aug · 22 pages, no figures. Substantially revised and expanded version. The one-sided condition on the second characteristic component has been removed. Exact small-data lifespan asymptotics, a global-existence alternative and long-time asymptotics are established. The results are also extended to endpoint exponents and to a class of general nonlinear sources. Title changed

Group Theory

Rota—Baxter operators on extraspecial groups

Theory of Rota–Baxter operators on rings and algebras has been developed since 1960. In 2021, L. Guo, H. Lang, Y. Sheng have defined the notion of Rota—Baxter operator on a group. We classify all Rota—Baxter operators on the finite Heisenberg group and the extraspecial group of order $p^3$. We also investigate an analogue of the spectral property for Rota—Baxter operators on algebras.
Read the full abstract

Solar and Stellar Astrophysics

Probing magnetic fields in stars: A 2D oscillation framework including rotation and magnetism

Understanding the role of internal magnetic fields in stars remains a major challenge for the description of angular momentum transport and stellar evolution. It is therefore essential to probe these magnetic fields within the star, and asteroseismology provides a powerful means to do so. In this work, we present a new implementation of the 2D oscillation code ( ), which incorporates the effects of both stellar rotation and magnetic fields. Adaptive Code of Oscillations towards Realistic modeling The code has been rendered modular, making it possible to specify the set of equations to solve and the assumptions through a symbolic calculus approach. The full set of adiabatic, non-radial pulsation equations is solved using a spectral approach for the angular part of the modes and high-order finite differences for the radial part. As a first step, we focus on a magnetic field that is purely toroidal and axisymmetric about the star's rotation axis. The numerical results are compared against first-order perturbative predictions in the weak-field regime and with the traditional approximation of rotation and magnetism (TARM) in the case of stronger magnetic fields. We validate this new implementation of against perturbative and TARM approaches, showing good agreement and demonstrating the robustness of the code. The breakdown of these methods provides general validity limits. We show that internal magnetic fields leave signatures in the period spacing of g modes. These features provide a promising seismic diagnostic to probe deep stellar magnetism in {\gamma} Dor stars. Future work will aim to extend this framework to more realistic magnetic field topologies and a broader range of pulsating stars, including red giants.
Read the full abstract

All authors Antoine Fort, Rhita-Maria Ouazzani, Lucas Barrault, Louis Manchon, Ludovic Petitdemange

Statistical Mechanics

A Generalized Model for Disordered Random Sequential Adsorption with Charge-Dependent Deposition

We generalize a one-dimensional random sequential adsorption model (RSA) of charged unit segments in which the deposition position follows a beta kernel determined by the charges bounding the available gap and those of the incoming particle. The interaction parameter \(0\leq\lambda<1\) interpolates between uniform car parking and strongly localized deposition. We derive a four component generating function equation that provides exact recursions for the complete occupation statistics, although only the mean and variance are analyzed in detail here. Exact finite-shell solutions serve as benchmarks for the Gauss–Jacobi quadrature and Monte Carlo simulations. We prove that, although boundary charges affect finite size behavior, the four boundary states have the same asymptotic mean and variance densities, consequently, we prove that the jammed density is self-averaging. Numerical results show that charge selectivity increases the coverage, especially for mixtures dominated by opposite endpoint charges, which also exhibit strongly reduced fluctuations near \(\lambda=1\). The model provides a tractable connection between uniform RSA, dynamically generated disorder, and interaction-driven deposition.
Read the full abstract

Computer Vision

GeoFF3D: Coordinate-Anchored Feed-Forward Reconstruction for Large-Scale UAV Mapping

Existing feed-forward 3D reconstruction methods typically process a bounded number of images and recover cameras and geometry in local or internally normalized frames. Extending them to large-scale UAV mapping requires scalable multi-chunk processing and reliable aggregation, while full Sim(3) alignment can become unstable for near collinear trajectories. We present GeoFF3D, which combines a coordinate-anchored model with a spatial large-scale reconstruction framework (SLRF). The model uses georeferenced camera translations and optional geometric priors to predict camera poses and dense point maps directly in a gravity-aligned Z-up metric frame. SLRF partitions images into spatially overlapping chunks, propagates shared-view priors, and aggregates local reconstructions hierarchically, while remaining applicable to different bounded-view models. Across nine aerial mapping blocks, GeoFF3D achieves the best average reconstruction quality, improving F@5 from 0.829 for Pi3X + SLRF to 0.877. On long UAVScenes sequences, it reaches 0.848, compared with 0.687 for Pi3X + SLRF and 0.451 for the strongest evaluated SLAM/streaming baseline. GeoFF3D reconstructs 2,000 images in approximately five minutes, demonstrating scalable and robust large-scale UAV reconstruction.The code is available at https://github.com/yanxian-ll/GeoFF3D.
Read the full abstract

All authors Xiang Yang, Yongli Wang, Yunsheng Zhang, Jun Li, Hao Chen, Haifeng Li

Instrumentation and Methods for Astrophysics

Development, Configuration and Performance Characterization of a Scalable BETA ASIC-Based Readout System for Multi-Channel SiPM Detectors

We present the design, configuration, and performance characterization of a scalable readout system based on the BETA application-specific integrated circuit (ASIC), developed to meet stringent requirements on noise, linearity, dynamic range, and power consumption for multi-channel silicon photomultiplier (SiPM) detectors in spaceborne instrumentation. The readout electronics consists of modular interface boards (FIBs) hosting multiple BETA ASICs and controlled by a field-programmable gate array (FPGA), which provides configuration, data acquisition, and global trigger generation. Multiple BETA FIBs were tested in a dedicated optical setup enabling simultaneous readout of a large number of channels. The system performance was evaluated using three S13552-10 SiPM arrays manufactured by Hamamatsu for the FIT detector, a scintillating-fiber tracker developed for charged cosmic-ray particle tracking and charge measurement in the HERD mission. We describe the configuration procedures and performance measurements of the readout system, including gain calibration, linearity characterization, and threshold response. In addition, we present the development of a global internal trigger logic for the identification of ionizing particles in the FIT detector. The results demonstrate the stability, scalability, and suitability of the developed BETA-based readout system for large-scale multi-channel SiPM detector applications in space experiments.
Read the full abstract

All authors Giulio Lucchetta, Keerthana Rajan Lathika, Benjamin Nobre Hauptmann, Philipp Azzarello, Oscar Blanch, Juan Boix, Laia Cardiel, Arturo Castaño Gallardo, Albert Espinya, Jennifer Maria Frieden, David Gascon, Daniel Gubermann, Kimberly Sarah Keyser, Manel Martinez, Joan Mauricio, Marina Orta-Terré, Chiara Perrina, Javier Rico, Andreu Sanuy, Monong Yu

Number Theory

Norm Principles for ${\rm Go}(q)$ and ${\rm Spin}(q)$

The aim of this article is to discuss the proof that the norm principle holds for the group of similitudes ${\rm Go}(q)$ and the spinor group ${\rm Spin}(q)$ for a quadratic form $q$ defined over $F$.

High Energy Physics (Phenomenology)

Searching for vectorlike $T$ quarks in the $T\to tZ$ channel at a future muon-proton collider

We investigate the discovery potential of a singly produced vector-like top quark $T$ in the $T\to tZ$ decay channel at future muon-proton ($\mu p$) colliders, considering three benchmark center-of-mass energies: $\sqrt{s}=5.29$, $6.48$, and $9.16$~TeV. The baseline signal final state consists of a semileptonic top decay $t\to bW\to b\ell\nu$ and a hadronic $Z\to q\bar{q}$ decay, where the collimated quark pair from the $Z$ boson is reconstructed as a single large-radius fat jet $J$. Full detector-level Monte Carlo simulations are performed for both the signal process and the dominant Standard Model backgrounds. We derive the $5\sigma$ discovery contours and the $95\%$ confidence level exclusion limits in the two-dimensional parameter space spanned by the VLQ mass $m_T$ and the effective coupling $g^*$. A comparison with the $T\to Wb$ channel and with searches at other collider facilities reveals that the $T\to tZ$ decay mode provides complementary sensitivity and offers unique advantages for probing high-mass vector-like $T$ quarks in the boosted jet topology.
Read the full abstract

High Energy Physics (Phenomenology)

Investigation of S-wave tetraquark bound and resonant states with all Jacobi coordinates

We systematically explore the $S$-wave tetraquark systems $Qs\bar{n}\bar{n}$, $QQ\bar{n}\bar{n}$, $QQ\bar{Q}\bar{Q}$, and $ss\bar{s}\bar{s}$ ($Q=c,b$; $n=u,d$) within the constituent quark potential model. We incorporate all K-type Jacobi coordinates in addition to the conventional H-type configurations, optimize the basis expansion via a stochastic parameter generation strategy, and apply the complex scaling method to identify bound and resonant states. Our calculations demonstrate that while conventional H-type configurations suffice for low-lying states such as the $T_{cc}(3875)^+$ molecular candidate, the inclusion of K-type configurations becomes important for extracting highly excited resonances, allowing higher-energy resonances absent in H-only calculations to be identified. Furthermore, we identify resonance candidates for the $T_{cs0}(2900)$, $X(6900)$, and $X(7200)$, whereas the absence of fully-strange compact poles below 2.6 GeV challenges the interpretation of $\phi(2170)$ and $X(2370)$ as $S$-wave compact $s s \bar{s} \bar{s}$ tetraquarks.
Read the full abstract

All authors Xin-He Zheng, Yao Ma, Liang-Zhen Wen, Shi-Lin Zhu

Computation and Language

Embedding Models for Stance-Aware Argument Retrieval

In computational argumentation, obtaining arguments that explicitly support or attack given claims is a critical precursor to downstream reasoning tasks. When these supporting and attacking arguments are to be retrieved using semantic search methods, they need to be assessed for topic-relevance to the claims of interest as well as for correctness of their (positive or negative) stance towards the claims. In this paper we explore how dense embedding models (hereafter, models), powering modern retrieval pipelines, can serve as the basis of semantic search incorporating this dual assessment. We show experimentally that existing models struggle with asymmetric reasoning, exhibiting a strong bias toward topical overlap while ignoring instructional stance. We also show that correcting this bias via contrastive training triggers a new failure mode where models over-correct, over-fixating on polarity keywords (e.g., "supports" or "refutes") at the expense of the semantic topic. We thus introduce diagnostic word-ablation metrics to quantify this phenomenon and propose a data-centric solution. By implementing a balanced argument curriculum alongside LLM-augmented, stance-inverted arguments, we force the (embedding) models to learn deeper directional logic rather than exploiting superficial lexical shortcuts. Our evaluation demonstrates that, for sufficiently powerful models, this approach can alleviate the observed overcorrection, achieving further improvements in stance-aware argument retrieval.
Read the full abstract

Combinatorics

Unimodular triangulations and Ehrhart theory for two families of Hermite normal form simplices

We study regular unimodular triangulations, the integer decomposition property, and Ehrhart-theoretic properties of two families of Hermite normal form simplices. We first consider the one-row case associated with the vector $(N - 1, \dots ,N - 1 , N)\in \mathbb{N}^d$, and completely characterize when the corresponding simplices admit a regular unimodular triangulation. Our constructions are explicit and also yield closed formulas for the $h^\ast$-polynomial and the local $h^\ast$-polynomial. Moreover, we prove Ehrhart positivity and derive explicit dimension-dependent conditions under which the Ehrhart polynomial is not unimodal. Finally, we extend our approach to the two-row cases associated with $(1, \dots ,1 , N)\in\mathbb{N}^d$ and $(M-1, \dots ,M-1, M, 0)\in\mathbb{N}^d$. In these cases, we construct regular unimodular triangulations, derive closed formulas for the $h^\ast$-polynomial and the local $h^\ast$-polynomial, and prove Ehrhart positivity.
Read the full abstract

Artificial Intelligence

LoopArena: Benchmarking Models as Runtime Controllers for Loop Engineering

Loop Engineering is emerging as a practice for organizing development work around coding agents. Instead of writing each prompt by hand, practitioners design loops that monitor progress, assign work, run checks, and decide what the agent should do next. Even with a capable coding agent, a loop may trust a stale progress note, skip needed verification, spend its budget in the wrong direction, or stop before the task is safe to submit. Yet the final outcome of one end-to-end run cannot tell whether success or failure reflects the loop's guidance or the coding agent's ability to carry out the task. We introduce LoopArena, a benchmark for evaluating how well one model can guide a separate coding agent through a long-running task. The model under evaluation is the Controller: after each coding round, it receives a structured summary of the run and instructs a separate, fixed coding agent, the Worker, on what to do or verify next, or decides whether to stop. LoopArena evaluates this ability in three complementary settings that differ in execution scope and cost. Type I consists of 160 low-cost control questions; its execution cost is incurred once during benchmark construction, so each new Controller is evaluated through a single four-way choice with no code execution. Type II evaluates Controller guidance on 80 selected task slices, while Type III evaluates the corresponding 80 full tasks from their original states. On full tasks, the best observed Strict Success Rate is 36.25%, leaving substantial room for improvement in long-horizon loop control. Across Controllers, the paired reduction in estimated inference cost averages 60.6%, and Type II produces a positively correlated ordering under the main Core criterion (Spearman's rho = 0.7353). We release the benchmark data and evaluation code at https://github.com/AMAP-ML/LoopArena.
Read the full abstract

All authors Yi Wang, Haopeng Zhang, Chengxiang Huang, Rui Dai, Kaikui Liu, Piotr Koniusz, Xiangxiang Chu

Fluid Dynamics

Unsteady airfoil aerodynamics in attached flow: From unsteady thin airfoil theory to wind turbine application

Attached-flow unsteady aerodynamics underpin the dynamic stall models used in wind turbine aerodynamic and aeroelastic codes, particularly over the outboard blade region that dominates power production and loading. Here, angle of attack and relative velocity vary while the flow remains predominantly attached. A complete engineering model combines airfoil polar lookup, shed-wake memory, non-circulatory loads, consistent force definitions, and separated-flow dynamics. Although the underlying theory is classical, existing descriptions do not provide a complete and internally consistent implementation route for these attached-flow contributions in wind turbine solvers. This work formulates the attached-flow contributions as lift, drag, and moment coefficients for blade-element momentum (BEM), lifting-line (LL), and actuator-line (AL) methods, for use with 2-D airfoil polars from measurements or CFD. Starting from classical unsteady thin-airfoil theory, circulatory and non-circulatory loads are derived in coefficient form and key modeling choices are clarified. Shed-wake memory is formulated using downwash velocity rather than angle of attack as the aerodynamic state variable. Rotor-level implementation is verified using two cases. A coned straight-blade case provides a cross-method benchmark and quantifies thrust and power errors caused by omitting three required contributions. Lift-direction projection has the largest influence on power, while omitting the mid-chord heaving-acceleration term removes the non-circulatory normal-force cancellation and produces a thrust error. A zero-onset-flow vertical-axis wind turbine (VAWT) case verifies that, in the ideal thin-airfoil limit, all circulatory and non-circulatory contributions cancel, yielding zero total rotor torque.
Read the full abstract

Robotics

STEGNav: Spatio-Temporal Event Graph Reasoning for Multimodal Lifelong Object Navigation

Multimodal lifelong navigation requires an agent to autonomously explore unseen environments while sequentially completing navigation tasks specified by object categories, language descriptions, or reference images. Existing methods primarily accomplish these tasks by constructing state-centric semantic scene graphs. By treating scene graphs as persistent repositories of semantic observations, these methods struggle to distinguish similar instances, jointly represent semantic targets and exploration frontiers, and effectively exploit navigation memory and trajectory experience. To address these limitations, we propose Spatio-Temporal Event Graph Navigation (STEGNav), a training-free framework that extends conventional scene graphs into spatio-temporal event graphs along complementary spatial and temporal axes. The spatial axis performs query-conditioned instance grounding and jointly represents semantic targets and occupancy-aware exploration frontiers characterized by reachability, path cost, and exploration utility. The temporal axis employs trajectory-aware dual-window memory to retain recent decision–trajectory events and verified cross-subtask navigation outcomes. A VLM-based navigation agent reasons over the resulting spatio-temporal event graph and selects either a target instance or an exploration frontier as its next navigation goal. STEGNav achieves 66.3% SR and 39.7 SPL on GOAT-Bench, as well as SR scores of 64.0% and 69.4% on HM3Dv1 and HM3Dv2, respectively. Ablation studies and error analyses validate the complementary effects of the two axes, demonstrating that event-driven spatio-temporal representations improve navigation reliability and cross-subtask experience reuse.
Read the full abstract

All authors Yang Chen, Zhenyu Huang, Wenbo Fu, Danyang Peng, Shi-Yu Tian, Kun-Yang Yu, Lan-Zhe Guo

Differential Geometry

Reduction of symmetric time-dependent Hamiltonian systems I: presymplectic principal $\mathbb{R}$-bundles

The reduction of mechanical presymplectic structures and its application to the reduction of time-dependent Hamiltonian systems were developed in a recent paper (Gutiérrez Sagredo et al, 2025). This approach overcomes some limitations of the original Albert reduction (Albert, 1989). In this paper we will describe the reduction of time-dependent Hamiltonian systems using the extended formalism. This process combines cotangent bundle reduction with reduction of presymplectic structures of corank 1 and 2. These results generalize previous work (Lacirasella et al, 2012), in the same direction. Several examples will be presented to illustrate the theory from this new perspective.
Read the full abstract

All authors C. Benítez, D. Iglesias Ponte, J. C. Marrero, E. Padrón

High Energy Physics (Phenomenology)

Color screening versus thermal decay as the mechanism of $\Upsilon$ suppression in high energy nuclear collisions

To clearly identify the mechanism behind the suppression of heavy quarkonium in relativistic heavy-ion collisions, we study $\Upsilon$ production at RHIC energies by solving its transport equation driven solely by the suppression rates. By calculating the nuclear modification factor and comparing it with experimental data, we find that the sudden suppression governed by the color-screening temperature cannot simultaneously describe both the ground and excited states of the $\Upsilon$, whereas the continuous suppression induced by thermal decay successfully reproduces all the $\Upsilon$ measurements. This provides strong evidence that inelastic scatterings with thermal partons, rather than color screening, dominate quarkonium suppression in heavy-ion collisions.
Read the full abstract

All authors Yida Yang, Baoyi Chen, Jiaxing Zhao, Pengfei Zhuang

Machine Learning

Parser Transitions Already Reveal Structural Responsibility: KV Persistence for Structured Generation

LLM agents use structured generation for JSON, SQL, and function calls. Constrained decoding ensures valid formats, but KV compression can discard history needed for correct function and argument selection. Structural responsibility can diverge from model saliency: importance scores can overlook generated tokens' structural roles. Parser transitions already computed for legality encode these roles, including required arguments and structural boundaries, providing a free retention signal. We introduce PASK (Parser-Aware Structural KV Persistence), which reuses this signal for KV persistence by layer group without online importance estimation. Task-error sensitivity sets protection floors, while attention-output distortion ranks upgrades under the remaining budget. Offline calibration produces a frozen table for structural lookup and budget accounting. PASK outperforms budget-controlled baselines on function calling and text-to-SQL, improving BFCL accuracy by up to 8.67 pp over the strongest comparator. Against Full KV, it cuts peak GPU memory by 47.5 percent and achieves up to 3.3 times TPOT speedup.
Read the full abstract

Instrumentation and Detectors

Full-field fluorescence computed tomography (F3CT) using a calibrated virtual cone-beam pinhole geometry

We present F3CT, a synchrotron-based hyperspectral full-field fluorescence computed tomography technique that avoids raster scanning by combining a pinhole aperture with an energy-resolving 2D detector. A virtual cone-beam model and two-stage calibration-reconstruction workflow enable 3D elemental mapping under full-field illumination. The method is demonstrated on biological and geological specimens, resolving silver-stain distributions in zebrafish tissue and high-energy fluorescence signatures in rock cores. Phase-contrast tomograms acquired sequentially under the same experimental geometry provide co-registered structural context. F3CT provides a high-throughput route to 3D XRF imaging and establishes a foundation for correlated structural and chemical tomography, with potential for future in-situ and operando implementations.
Read the full abstract

All authors Thomas Zillhardt, Yunhui Chen, Alexander Rack, Matthew Veale, Matt Wilson, Philip J. Withers, Nicola Viganò

Machine Learning

An algebraic proof of Colombo's difference-power determinant conjecture

Let $n\ge2$ be even, let $\lambda=(\lambda_1,\ldots,\lambda_n)\in\mathbb{R}^n$ have pairwise distinct coordinates, and define the difference-power matrix \[ A_d(\lambda) := \bigl[(\lambda_r-\lambda_s)^d\bigr]_{r,s=1}^n, \qquad d\in\mathbb{N}. \] In 1928, Colombo proved that $\det A_{n-1}(\lambda)\ne0$—and hence $\det A_{n-1}(\lambda)>0$—and that $\operatorname{rank} A_d(\lambda)=d+1$ for $0\le d<n-1$. He conjectured that \[ \det A_d(\lambda)\ne0 \qquad\text{for every } d\ge n-1. \] For even $d$, the conjectured nonsingularity follows from previously published results on distance-power matrices. The remaining open cases were therefore the supercritical odd exponents $d\ge n+1$. We prove nonsingularity for all these odd exponents, thereby completing Colombo's conjecture. Consequently, \[ \operatorname{rank} A_d(\lambda)=\min\{n,d+1\} \qquad(d\in\mathbb{N}). \] Our proof converts a hypothetical kernel vector into a real binary form having more projective real linear factors, counted with multiplicity, than its real Waring length permits.
Read the full abstract

All authors Kun Li, Li Tie, Peng Wang, Zihan Liu

Machine Learning

Learning to Transfer Across Modes: Towards Unified Urban Mobility Forecasting

Urban transportation systems consist of multiple mobility modes that coexist within the same city and exhibit complex interdependencies, leading to correlated demand dynamics across modes. However, forecasting demand jointly across different modes remains challenging due to substantial heterogeneity in space and the limited availability of historical data for emerging modes. Existing forecasting methods are largely developed for individual mobility modes and implicitly assume compatible spatial structures between source and target systems, which severely restricts their applicability in multi-modal settings. To address these challenges, we propose TransMod, a unified framework for urban mobility demand forecasting that enables effective knowledge transfer across heterogeneous mobility modes. TransMod constructs a shared zone-level spatial representation that aligns mobility systems with different spatial granularities into a common space, thereby reducing structural mismatch and distributional shift. Built on this unified representation, TransMod further learns transferable spatio-temporal patterns from data-rich source modes and adapts them to data-scarce target modes, alleviating the dependence on extensive target-domain histories. Extensive experiments on real-world datasets demonstrate that TransMod consistently outperforms existing approaches and provides robust forecasting performance under limited target data.
Read the full abstract

Computer Vision

Non-Uniform Quantisation for 3DGS Compression

3D Gaussian Splatting (3DGS) has emerged as a powerful technique for novel view synthesis, yet its high bitrate requirements pose significant challenges for storage and transmission. To enable practical applications and ensure interoperability within the 3DGS ecosystem, standardised compression formats are essential. In this paper, we propose a novel non-uniform quantisation scheme specifically tailored for 3DGS models. Our approach adapts to the underlying data distribution by applying importance-weighted quantisation and eliminating post-voxelisation redundancy through importance weighted merging. Extensive evaluations on benchmark datasets demonstrate that our method achieves state-of-the-art compression performance. Furthermore, the proposed scheme is compatible with any point-cloud-based representation and is intended as a formal contribution to the upcoming MPEG 3DGS compression standardisation activities.
Read the full abstract

Artificial Intelligence

RECAST: Recent & Context-Aware Sampling for Test-Time Adaptation in Streaming Biosignals

Streaming biosignals vary across subjects and drift over time, so population-trained models lose accuracy during long-term monitoring. Test-time adaptation (TTA) enables online personalization by updating the model on incoming samples. But in a stream, a basic question is left open: which samples should drive each update? Using all buffered samples blurs the update with irrelevant segments. Using only the latest segment makes the update noisy and unstable. The most useful samples are recent, aligned with the current physiological state, and reliable enough to learn from. We propose RECAST (REcent & Context-Aware Sampling for TTA), a lightweight sampling module for buffered TTA frameworks. RECAST builds each adaptation batch from three signals: temporal recency, contextual similarity, and predictive reliability. It changes only which samples are used, leaving the model and the training objective unchanged. On two blood-pressure datasets, RECAST improves estimation accuracy and trend tracking over baselines and ablations. The per-patient gains are statistically significant on both datasets, with broad improvement on the regular benchmark and gains concentrated on the hardest patients in the emergency-department setting. RECAST stays practical, adding only sub-second latency per segment on a single GPU and CPU core.
Read the full abstract

Robotics

Spatial-Semantic Reasoning using Large Language Models for Efficient UAV Search Operations

We present a real-time semantic navigation framework for Unmanned Aerial Vehicles (UAVs) focused on improving time efficiency in the Object Goal Navigation (ObjectNav) task. Central to our approach is a Large Language Model (LLM) that interprets user-provided natural language instructions and performs semantic reasoning over detected objects and spatial context to prioritize high-probability search regions. The system combines real-time object detection, 3D spatial mapping, and polynomial spline interpolation for smooth and feasible UAV trajectory planning. Unlike prior methods that rely on offline reasoning or simulator-constrained action spaces, our framework can operate in real time, continuously updating semantic relevance based on new observations. Experiments in both simulated and real-world settings demonstrate reductions in mission duration while maintaining high search accuracy, underscoring the effectiveness of LLM-guided reasoning for time- efficient UAV-based ObjectNav.
Read the full abstract

All authors Marin Maletic, Marijana Peti, Tamara Petrovic, Stjepan Bogdan

Materials Science

Phenomenological Growth Regimes in Liquid-Precursor CVD of MoS$_2$ on Functional Substrates

The integration of two-dimensional transition-metal dichalcogenides (TMDCs) onto functional substrates remains constrained by stochastic vapor-phase growth dynamics. Here, we show that liquid-phase precursor chemical vapor deposition (CVD) of MoS$_2$ introduces growth conditions that are consistent with a substrate-influenced reaction-diffusion process. By utilizing pre-growth spin-coated MoO$_3$ intermediates across a diverse crystalline library (sapphire, SrTiO$_3$, rutile TiO$_2$, MgO, and 6H-SiC), we find that substrate-dependent variations in precursor wetting, surface chemistry, and inferred mass-transport constraints correlate with distinct growth morphologies. These substrate-dependent growth regimes are interpreted in terms of reduced effective lateral growth length on SrTiO$_3$, possible precursor anchoring on TiO$_2$, likely chemical surface restructuring on MgO, and possible step-edge growth on SiC. Raman and photoluminescence spectroscopy reveal substrate-dependent variations in vibrational and optical response that correlate with differences in strain, charge environment, and dielectric screening. Ultimately, this work highlights a substrate-dependent reaction-diffusion framework as a potentially useful route for tuning the structural and optical properties of large-area 2D materials.
Read the full abstract

All authors Osamah Kharsah, Kilian Mouchel, Yossarian Liebsch, Cathy Sulaiman, Joel Verlande, Anke Hierzenberger, Abdallah Alghazali, Clara Grygiel, Stéphane Guillous, Henning Lebius, Marika Schleberger