Mesoscale and Nanoscale Physics
Multiscale Modeling of Ion Transport in Nanopores: Fitting Implicit-Water Radial Diffusion Profiles to Explicit-Water Molecular Dynamics
Mónika Valiskó, Salman Shabbir, Eszter Molnárné Lakics et al. (+2 more)
We develop a multiscale approach for incorporating molecular-scale transport information into computationally efficient models of ion transport through nanopores. A radially varying effective
diffusion coefficient profile is fitted to radial conductivity profiles obtained from explicit-water molecular dynamics (MD) simulations. The fitting is performed within the NP+LEMC framework, which combines the Nernst–Planck equation with Local
Equilibrium Monte Carlo to account for ion correlations beyond mean-field approximation. We apply the approach to NaCl, CaCl$_2$, and their mixtures in a negatively charged silica nanopore. The resulting
diffusion coefficient profiles reproduce the radial current distributions of the MD simulations, including the strong suppression of ionic mobility near the pore wall that cannot be captured by a spatially constant
diffusion coefficient inside the pore. NaCl and CaCl$_2$ exhibit qualitatively different transport behavior: the former is cation selective due to enhanced near-wall Na$^+$ conduction, whereas the latter shows weak anion selectivity because strongly bound Ca$^{2+}$ ions have strongly suppressed mobility near the surface. For NaCl–CaCl$_2$ mixtures, preferential Ca$^{2+}$ binding leads to nonlinear changes in ionic conductance. The approach establishes a bridge between experimentally relevant device behavior and computationally efficient reduced models, with explicit-water MD providing the molecular-scale information that can be incorporated into the effective transport coefficients.
Read the full abstract ▾
All authors Mónika Valiskó, Salman Shabbir, Eszter Molnárné Lakics, Zoltán Ható, Dezső Boda
Machine Learning (Statistics)
PACE: Plug-and-Play Contextual Embedding for Feature Screening with Pretrained Tabular Foundation Models
Qi Qin, Erbo Li, Ting Wei et al. (+4 more)
In high-dimensional tabular learning, feature screening provides a lightweight, model-agnostic way to remove irrelevant features before model fitting. However, scoring raw values directly can miss nonlinear or distributional structure. We introduce PACE (Plug-and-Play Contextual
Embedding), which inserts a frozen tabular foundation model (TFM) column encoder before an existing feature-scoring rule, expanding each feature into a higher-dimensional contextual representation. Across controlled studies, PACE improves raw-space screening of complex nonlinear dependence with only modest additional encoding cost. These gains translate to downstream prediction on TALENT
datasets: PACE-DC improves binary AUC by 0.077 and multiclass macro-AUC by 0.064, with a median normalized RMSE improvement of 0.063 across ten learners. Matched random-weight and random-feature controls show that PACE gains from pretrained structure beyond generic dimensional expansion. PACE further achieves favorable performance–time trade-offs against task-fitted selectors and attribution-based methods, positioning pretrained column geometry as a reusable upstream primitive for tabular learning.
Read the full abstract ▾
All authors Qi Qin, Erbo Li, Ting Wei, Zizhou Huang, Zixuan Qin, Wu Wang, Yifan Sun
Computation and Language
Error-Supervised Synthetic Learner Writing for Automated Essay Scoring
Duy Anh Nguyen
Synthetic essays can help reduce dependence on human-written data in Automated Essay Scoring (AES). However, they often lack realistic errors, limiting their ability to represent authentic human writing, particularly when the target texts are intended to resemble those produced by
language learners. In this study, we present a simple approach that introduces error supervision into synthetic essay generation. Specifically, we fine-tune an LLM generator on error-annotated texts of the kind commonly used in Grammatical Error Detection (GED). To assess the utility of the proposed approach, we fine-tune and evaluate AES scorers under three data conditions: authentic essays, synthetic essays generated conventionally, and synthetic essays generated using our proposed approach. The results show that in the larger-data settings, the proposed approach outperforms the conventional synthetic baseline in 11 out of 12
dataset-metric comparisons, with performance in some cases approaching that of models trained on authentic essays. Despite these gains, performance under extremely low-resource settings remains mixed, with advantages over the conventional baseline only becoming more apparent at 200 training essays, although not consistently across
datasets. Qualitative and quantitative analyses further show that the proposed approach produces learner-like errors whose distributions broadly resemble those observed in authentic essays.
Read the full abstract ▾
Complex Variables
Sampling and Interpolation in Gaussian Mixed-Norm Fock Spaces
Xiang Fang and Pham Trong Tien
We establish sharp sampling and interpolation criteria for the radial–angular Gaussian mixed-norm Fock spaces $\mathcal F_\alpha^{p,q}$, where $\alpha>0$ and $0<p,q\le\infty$. The natural point-evaluation scale leads to the normalized restriction $R_{\alpha,\Lambda}^{p,q}f = \bigl(f(\lambda)e^{-\frac{\alpha}{2}|\lambda|^2}(1+|\lambda|)^{\frac{1}{q}-\frac{1}{p}}\bigr)_{\lambda\in\Lambda}$. With this normalization, the critical density is the classical value $\frac{\alpha}{\pi}$ throughout the
Banach and quasi-
Banach ranges. For an arbitrary locally finite set $\Lambda$, sampling is equivalent to $D_{\mathrm{sep}}^{-}(\Lambda)>\frac{\alpha}{\pi}$, together with relative separation when $p<\infty$; no relative-separation condition is required when $p=\infty$. Here $D_{\mathrm{sep}}^{-}$ is the supremum of the lower Beurling densities over separated subsets of $\Lambda$. Interpolation is equivalent to separation and $D^{+}(\Lambda)<\frac{\alpha}{\pi}$. The same criterion characterizes interpolation for the little endpoint $f_\alpha^{p,\infty}$, and in both settings the normalized restriction map admits a bounded linear right inverse. The sufficiency arguments are based on rapidly localized
Hilbert dual and Lagrange atoms together with a weighted localized synthesis
theorem on the full mixed-norm scale. For necessity, we prove that lower stability of a rapidly localized
matrix on a weighted annular mixed sequence space implies lower stability on $\ell^2$. The proof first passes to $\ell^\infty$ by translated
polynomial cutoffs and a commutator estimate, and then uses the $p$-independence
theorem for the Sjöstrand class. Applied through a fixed
Hilbert sampling
lattice, this reduces the strict density conditions to the classical
Hilbert Fock sampling and interpolation
theorems.
Read the full abstract ▾
Mathematical Finance
Extremal Mean-Variance Functionals over Wasserstein Balls: Applications to Risk Sharing
Wenjun Jiang, Yiying Zhang and Zhenfeng Zou
We characterize the worst- and best-case values of a mean-variance functional over a 2-Wasserstein ball. Using
quantile representations and the geometry of attainable means and standard deviations, we reduce both infinite-dimensional problems to scalar equations and construct the extremal laws as location-scale transformations of the reference distribution. Their values depend on the reference law only through its first two moments. We then derive dual representations and formulate proportional risk sharing under heterogeneous beliefs as a finite-dimensional optimization problem. Under homogeneous beliefs, we show that the classical proportional allocation remains optimal for every ambiguity radius and $\alpha$-maxmin weight. Finally, we study a coupled distortion-variance functional and characterize its worst-case
quantile through a convex-envelope construction, allowing the extremal law to change in shape as well as location and scale.
Read the full abstract ▾
Software Engineering
VibeMemBench: Evaluating Memory Systems for Coding Agents on Real Repository Coding Tasks
Liyang Fan, Yingcheng Shi, Yongbin Li et al. (+7 more)
Coding
agents operate on real repository coding tasks, and persistent
memory systems promise to reuse experience across tasks. Yet existing evaluations do not show whether those systems improve executable repository work. Repository
benchmarks test code changes but do not isolate
memory, while
memory benchmarks score recall without measuring downstream coding outcomes. We introduce VibeMemBench, a
benchmark for evaluating
memory systems on 111 coding targets from 90 SWE-rebench V2 repositories and 3,634 history trajectories from the target repositories. The targets follow the SWE
benchmark style and cover bug fixes, feature requests, interface changes, and configuration work. An
agent edits each target codebase under a declared
memory condition. Executable tests decide task resolution. Each target is retained only when injected history experience improves its executable outcome in a reference setting, so every target carries a prior experience whose usefulness is verified by execution in that setting. The frozen verified experience is then transferred to five held-out solvers. Direct injection raises observed task resolution on four of them by 1.1 to 4.5 percentage points while lowering
agent steps on all five. Yet when four existing
memory systems must construct and retrieve experience from the same history, eleven of twelve solver and system pairings fail to exceed the matched
memory-off baseline. VibeMemBench exposes the gap between the useful experience that repository history holds and the experience existing
memory systems deliver for repository coding tasks.
Read the full abstract ▾
All authors Liyang Fan, Yingcheng Shi, Yongbin Li, Chenghao Sun, Xin Chen, Xander Xu, Hu Wei, Shiwen Ni, Min Yang, Jieping Ye
Computational Geometry
Bichromatic Line-Centers for Point Pairs
Jaegun Lee, Youjung Bae, Taehoon Ahn et al. (+2 more)
We study the bichromatic line-center problem for $n$ pairs of points in the plane. A feasible solution assigns one point from each pair to the red set $R$ and the other to the blue set $B$. The goal is to minimize $\max\{w^\circ(R),\,w^\circ(B)\}$, where $w^\circ(X)$ denotes the minimum width of a strip enclosing $X$; the midlines of the corresponding optimal strips define the line-centers of $R$ and $B$. We consider several variants induced by orientational constraints on line-centers and provide efficient
algorithms for each. For one line-center, which consists of computing a minimum-width strip that contains at least one point from each pair, we give an $O(n^2)$-time
algorithm. For two line-centers, we obtain an $O(n)$-time
algorithm when both are horizontal, and $\Theta(n\log n)$-time
algorithms when the two centers are parallel or when both orientations are prescribed. When exactly one orientation is prescribed, we give an $O(n^2)$-time
algorithm. Finally, for the unrestricted case, we present an $O(n^3\log n)$-time
algorithm.
Read the full abstract ▾
All authors Jaegun Lee, Youjung Bae, Taehoon Ahn, Sang Won Bae, Hee-Kap Ahn
Signal Processing
Fiber-Topology-Assisted Indoor Positioning Under UWB Degradation with Fixed-Lag Factor Graphs
Zhiyang Xue, Yuhan Jiang, Ziang Chen et al. (+6 more)
Ultra-wideband (UWB) provides absolute spatial constraints for indoor positioning, but these constraints become insufficient when multiple
radio links are blocked or unavailable. A calibrated sensing fiber, in contrast, provides a fixed spatial
topology that can be activated by distributed acoustic sensing (DAS) events independently of
radio propagation. Motivated by this complementarity, we present FTALoc-FGO, a
topology-aware fixed-lag factor-graph framework for fiber-assisted localization under UWB degradation. Along-fiber DAS events are mapped through the calibrated
topology into anisotropic spatial observations, where a single fiber layer provides a directional constraint and two orthogonal layers provide complementary two-dimensional information. Fisher-information analysis characterizes this
topology-dependent information and motivates the directional and joint fiber factors used in the fixed-lag graph. Experiments combining measured DAS events with simulated UWB ranges under controlled degradation evaluate the method over 30 paired realizations. Under complete UWB loss, FTALoc-FGO reduces the causal outage RMSE by 32.6% and 66.4% for single- and dual-layer routes, respectively, relative to robust UWB-only FGO. These results demonstrate that calibrated fiber
topology can provide complementary spatial support when
radio-based observations become insufficient.
Read the full abstract ▾
All authors Zhiyang Xue, Yuhan Jiang, Ziang Chen, Zhongxing Tian, Zeyu Feng, Lei Wang, Huan Huang, Dongdong Zou, Yi Cai
Formal Languages and Automata
Contributions to the hierarchy of probabilistic languages
Lothar Sebastian Krapp and Remo Nitschke
We reconsider the theory of probabilistic formal
languages generated by n-gram models and by probabilistic context-free grammars (PCFGs). The expected hierarchy of probabilistic grammars is established by proving that every probabilistic
language generated by an n-gram model is also generated by some PCFG, while some probabilistic
languages generated by PCFGs cannot be generated by any $n$-gram model. We introduce the notion of fully connected PCFGs, namely PCFGs in Chomsky normal form where every production rule only involving non-terminals has non-zero probability. Our main result shows that any probabilistic
language generated by an $n$-gram model differs from any probabilistic
language generated by a fully connected PCFG. Therefore, the class of probabilistic
languages generated by $n$-gram models is not a subset of the class generated by fully connected PCFGs.
Read the full abstract ▾
Computer Vision
G6D: Geometric Learning-Free RGB-D 6D Pose Solver for Robotic Manipulation
Yixuan Liang, William Chen, Yunan Wang et al. (+4 more)
6D object pose estimation is fundamental to robotic manipulation and automation. Recent zero-shot methods have significantly improved generalization to unseen objects, but most still rely on large-scale pretrained models with substantial GPU computation and
memory demands. These requirements complicate deployment on robotic platforms where perception, planning, and control share limited computational resources, while learned intermediate representations offer limited
geometric interpretability for task-specific adaptation. To address these limitations, we propose G6D, a learning-free, geometry-driven RGB-D 6D pose solver. Given an RGB-D observation, an object instance mask, camera intrinsics, and a CAD model, G6D generates pose hypotheses through template-based
geometric matching and refines them using silhouette and depth consistency, forming a purely geometry-driven pose estimation paradigm. This paradigm requires neither pretrained visual models nor target-specific training and preserves interpretable
geometric representations throughout pose estimation. Moreover, adjustable hypothesis counts provide flexible accuracy-computation trade-offs, while a CPU-only configuration supports deployment without GPU resources. Experiments on LineMOD and five BOP19
datasets demonstrate advanced performance. Real-world pick-and-place experiments further demonstrate G6D's applicability to robotic manipulation. The complete project is publicly available at https://ai4control.github.io/G6D-Project-Page .
Read the full abstract ▾
All authors Yixuan Liang, William Chen, Yunan Wang, Jizhou Yan, Zhao Jin, Changling Liu, Chuxiong Hu
Computer Vision
MaskVLA: Visual Masking Against Trajectory Overfitting of Vision-Language-Action Model
Yuxuan Jiang, Jiaying Huang, Ge Wang et al. (+9 more)
Vision-
Language-Action (VLA) models integrate
vision-
language understanding with executable
robot actions, enabling end-to-end learning for
robot control. However, our empirical analysis reveals that existing models exhibit severe trajectory overfitting when finetuned on limited
datasets. To guide the model in effectively utilizing wrist camera information, we propose MaskVLA, a masking-based fine-tuning strategy. By randomly masking a small portion of the main camera's visual information, the model is guided to autonomously learn more fine-grained, task-relevant, and effective visual features. This process leads to the emergence of robust policies, thereby enhancing the model's capability to tackle complex manipulation tasks and improving its generalization performance. Our method has been comprehensively evaluated on RoboTwin 2.0, achieving an average success rate improvement of 23.2% and 16.8% compared to $\pi_0$ and OpenVLA-OFT, respectively. Furthermore, experiments on real-world ALOHA
robots also demonstrate the effectiveness of our approach.
Read the full abstract ▾
All authors Yuxuan Jiang, Jiaying Huang, Ge Wang, Shenhao Yan, Jiahao Yang, Chengsi Yao, Qi Liu, Qing Zhao, Shuguang Cui, Yiming Zhao, Yatong Han, Zhen Li
Materials Science
Dual Role of Mobile Interstitials in Defect Kinetics: From Retardation to Acceleration
Shihao Zhang, Shihao Zhu, Junping Du et al. (+3 more)
Mobile interstitial atoms redistribute while defects migrate, complicating prediction of defect kinetics and leading to contradictory reports of acceleration and retardation. We formulate defect activation in a grand-canonical ensemble at fixed interstitial chemical potential and define a pathway free-energy landscape $\Delta G(\Lambda;T,\mu)$ and the corresponding activation free energy $\Delta G^{\ddagger}$. We compute these quantities using both hyperplane-constrained thermodynamic integration via a mean force relation and a two-state chemical-potential integration, denoted TSCPI. The latter requires sampling only the initial and transition states; a single chemical-potential integration then yields $\Delta G^{\ddagger}(T,\mu)$ across a wide $\mu$ range, enabling rapid mapping over temperature and chemical potential. Applied to vacancy
diffusion in FCC, BCC, and HCP metals with H (including in plane and cross plane
diffusion in Zr) and in BCC W with He, both free energy routes agree with
diffusion coefficients from molecular dynamics mean square displacements. The resulting maps reveal regimes of suppression and enhancement, including a crossover from suppression to enhancement with increasing hydrogen concentration. A site occupancy analysis links barrier shifts to state dependent site spectrum changes and transferable interstitial interaction terms.
Read the full abstract ▾
All authors Shihao Zhang, Shihao Zhu, Junping Du, Shuhei Shinzato, Ju Li, Shigenobu Ogata
Materials Science
Deterministic control of antiferromagnetic domain walls by circular phonons
Vladislav Bilyk, Nikolai Khokhlov, Viktoriia Radovskaia et al. (+12 more)
In an antiferromagnet, areas with different orientations of the
antiferromagnetic Néel vector, which plays the role of the order parameter, can coexist, thus forming
antiferromagnetic domains separated by domain walls. This suggests that antiferromagnets can possess the functionalities of
magnetic storage media. Whether one can benefit from these functionalities crucially depends on the availability of efficient means for deterministic control of
antiferromagnetic domain walls. Here, we demonstrate an approach to deterministically control domain walls in the van der Waals antiferromagnet Ni-doped $MnPS_3$ via dynamic engineering of the
crystal lattice. By resonantly exciting a pair of nearly degenerate orthogonal infrared-active $A_u$ and $B_u$
phonon modes with circularly polarized mid-infrared light, we induce helicity-dependent reconfiguration of 180{\deg}
antiferromagnetic domains, providing evidence for a
phonon-induced effective field conjugate to the Néel order parameter. The domain kinetics exhibit a pronounced helicity asymmetry governed by the interplay between domain-wall elasticity and defect-mediated pinning, enabling small
phonon-driven perturbations to accumulate into stable domain transformations. Our results establish dynamic
lattice control and defect engineering as means for deterministic control of
antiferromagnetic order.
Read the full abstract ▾
All authors Vladislav Bilyk, Nikolai Khokhlov, Viktoriia Radovskaia, Peter Kim, Pietro Diona, Ravi Kaushik, Luca Maranzana, Takeshi Hayashida, Sergey Artyukhin, Edwin Hang Tong Teo, Apoorva Chaturvedi, Carl S. Davies, Andrei Kirilyuk, Alexey Kimel, Dmytro Afanasiev
Number Theory
Refined sum formulas for finite multiple zeta values of level two
Zhenlu Wang
In this paper, we establish two classes of refined sum formulas for finite multiple
zeta values of level two, a variant of finite multiple
zeta values recently introduced by M. Kaneko et al. One is an explicit formula for the sum of level-two finite multiple
zeta values over indices in which one prescribed component is odd and all the remaining components are even. The other is an Ohno–Zagier type relation for level-two finite multiple
zeta values, which gives a generating function for sums of fixed weight, depth and height in terms of depth-one values. As applications, we derive sum formulas for fixed weight and depth and for several special heights.
Read the full abstract ▾
Computer Vision
Transferring Visual Explanations: How Cross-Architecture Knowledge Distillation Affects Model Interpretability
Aleks Czufarow and Ihor Babin
Deploying efficient neural networks is essential in resource-constrained environments, yet compact models often sacrifice interpretability - a critical in safety-critical domains such as autonomous driving and medicine. This study investigates whether Knowledge
Distillation transfers the spatial feature attribution of a large teacher network to a compact student. To assess the influence of the KD scheme on interpretability, we distill a ResNet-152 teacher into a ResNet-34 student on ImageNet-1K across five configurations by systematically varying the
distillation temperature and soft-label loss weight. Models are evaluated on top-1 accuracy, along with two interpretability metrics: Relevance Mass Accuracy and Relevance Rank Accuracy. These metrics are computed via Grad-CAM heatmaps benchmarked against ground-truth object masks. Our results show that top-1 accuracy ranges from 71.6% to 74.0%. For Grad-CAM, RMA ranges from 7.7% to 9.7% and RRA from 7.3% to 10.1%; for Guided Grad-CAM, RMA ranges from 16.1% to 18.6% and RRA from 15.9% to 21.5%. Interpretability proves far more sensitive to the soft-label weight than to the temperature: keeping the student anchored to hard labels preserves both accuracy and coarse localization, whereas weighting the teacher heavily degrades both. Fine-grained attribution, however, fell below the undistilled baseline in every configuration tested, indicating that logit
distillation transmits where a model attends more readily than the pixel-level structure of that
attention. We evaluate 12 cross-architecture combinations of convolutional and
transformer-based models, revealing that the inheritance of fine-grained spatial
reasoning is fundamentally bottlenecked by the student's intrinsic structural biases. To our knowledge, this is the first application of this interpretability-aware evaluation framework - previously used for neural network
pruning - to KD.
Read the full abstract ▾
Representation Theory
Affine Harish-Chandra center in positive characteristic
Gurbir Dhillon and Ivan Losev
Let $G$ be a split reductive group defined over a field of characteristic bigger than the Coxeter numbers of its simple factors. We identify the Harish–Chandra centers of the associated Kac–Moody vertex
algebras and enveloping
algebras at noncritical levels with
algebras of functions on
moduli spaces of connections for the Langlands dual group $\check{G}$. Namely, given a noncritical level $\kappa$ for $G$, consider the associated Kac–Moody vertex
algebra $V_\kappa(\mathfrak{g})$ defined on a formal disc $\mathscr{D}$, and its Harish–Chandra center of arc group invariants $$V_\kappa(\mathfrak{g})^{\mathscr{J}G} \subset V_\kappa(\mathfrak{g}).$$ Write $\check{\kappa}$ for the dual level for $\check{G}$, $\check{\kappa}^p - \check{\kappa}$ for its image under the Artin–Schreier map, and $\operatorname{Op}_{\check{G}}(\mathscr{D}^{(1)})_{\check{\kappa}^p - \check{\kappa}}$ for the
moduli space of $(\check{\kappa}^p - \check{\kappa})$-opers on the Frobenius twisted disc $\mathscr{D}^{(1)}$. We establish a canonical isomorphism $$\operatorname{Spec} V_\kappa(\mathfrak{g})^{\mathscr{J} G} \simeq \operatorname{Op}_{\check{G}}(\mathscr{D}^{(1)})_{\check{\kappa}^p - \check{\kappa}}.$$ There is a similar identification for the Harish–Chandra center of the filtered complete enveloping
algebra at level $\kappa$, where instead of $\mathscr{D}$ we need to consider the punctured disc $\mathscr{D}^\times$. The presence of these Harish–Chandra centers is a genuinely new phenomenon for loop groups at noncritical level in positive characteristic: these centers are nontrivial, unlike for loop groups at noncritical level in characteristic zero, and in particular are not the reductions mod $p$ of the characteristic zero Harish–Chandra centers, unlike for finite dimensional reductive groups or loop groups at critical level.
Read the full abstract ▾
Quantum Algebra
Grothendieck rings of strictly weakly integral fusion categories of rank six
Kai Wang, Jingcheng Dong and Libin Li
We classify, up to based-ring isomorphism, the Grothendieck rings of strictly weakly integral fusion categories of rank six over $\mathbb C$. The dimension grading and the classification of integral Grothendieck rings of rank at most five reduce the problem to twelve candidate fusion rings. We exclude four of them using the twist trace and the induction pairing in the Drinfeld center. The remaining eight admit known categorifications. In particular, every such category has commutative Grothendieck ring and Frobenius–Perron dimension in $\{8,10,16,20\}$.
Read the full abstract ▾
Probability
Operator-norm Sudakov minoration for Gaussian chaos of order two
Witold Bednorz, Rafał Martynek and Rafał Meller
We prove that an operator-norm separated family of
matrices satisfies $\mathbb{E}\sup_{A\in T} G^{T}AG' \geq ca\log |T|$, where G,G' are independent standard Gaussian vectors and a is the separation. The main information estimate concerns arbitrary separated coisometries: conditional
entropy is bounded by a source-dependent operator energy times $\log|T|$, up to an additive quadratic term in the common row dimension. An adaptive Gaussian experiment proves this estimate by charging actual information increments to one weighted posterior-
entropy potential. Convex separation and a Gaussian covering estimate then yield a bounded-radius result. To reach the general case, we first choose an operator scale preserving the Sudakov ratio, apply the known
Hilbert-Schmidt minoration, and recompute a common Gaussian block compression at the retained
entropy. This ordering preserves the normalization needed by the coisometry argument.
Read the full abstract ▾
Optimization and Control
Optimal High-Order Methods for Solving Monotone Variational Inequalities
Xinliang Zhang, Lesi Chen, Linxuan Pan et al. (+3 more)
We study second- and higher-order methods for solving smooth monotone variational inequalities (MVI). Monteiro and Svaiter (SIAM J. Optim., 2012) showed that a second-order method, NPE, converges at a rate of $\mathcal{O}(T^{-1.5})$. For convex-concave minimax optimization, a subclass of MVI problems, Chen, Liu, Luo, and Zhang (COLT 2025) recently improved this rate to $\tilde{\mathcal{O}}( T^{-1.75})$. However, the result has a substantial gap compared to the lower bound of $\Omega(T^{-2.5})$ established by Chen et al. (2026). In this paper, we propose a novel second-order method that achieves the optimal rate of $\mathcal{O}(T^{-2.5})$. Our
algorithm also extends to higher-order methods: for any integer $ p \ge 1$, we obtain a $p$th-order method with a convergence rate of $\mathcal{O}(T^{-(3p-1)/2})$, matching the known lower bounds and therefore establishing optimal rates across all orders.
Read the full abstract ▾
All authors Xinliang Zhang, Lesi Chen, Linxuan Pan, Chengchang Liu, Junchi Yang, Jingzhao Zhang
High Energy Physics (Theory)
Gauge Origami of the Conifold $\times$ $\mathbb{C}$
Taro Kimura and Go Noshita
We construct gauge-origami systems on the toric Calabi–Yau fourfold $\mathcal{C}\times\mathbb{C}$ from its brane brick model. Brick matchings and phase boundaries determine elementary D6- and D4-brane framings, respectively, realizing the magnificent-four, tetrahedron-instanton, and spiked-instanton sectors in a unified framed quiver
quantum mechanics. We define their partition functions by Jeffrey–Kirwan residues and describe the fixed points in terms of colored BPS
crystals. General gauge-origami configurations are obtained by combining these elementary sectors through their common fractional D0-brane quiver. This construction provides a first step toward gauge origami on general toric Calabi–Yau fourfolds.
Read the full abstract ▾
Image and Video Processing
Resolution-Flexible Decoding for Hybrid Neural Video Representations
Taiga Hayami, Masaya Takabe and Hiroshi Watanabe
Neural
video representations (NVRs) represent
videos using neural network parameters and, in hybrid formulations, frame-wise latent
embeddings. In conventional hybrid NVRs, the spatial size of frame-wise latent
embeddings and the decoder upsampling schedule depend on the target resolution, potentially requiring large and non-uniform upsampling factors. In this paper, we propose a resolution-flexible decoder framework for hybrid NVRs. The decoder is constructed from uniform \(2\times\) upsampling stages, whose target feature sizes are obtained by tracing the spatial resolution backward from the final output resolution. After each upsampling stage, the feature map is aligned with the target size by minimal padding or cropping when necessary. To support this progressive decoding process, we further use intermediate reconstruction supervision and a reconstruction-difficulty-aware frame sampling strategy based on recent frame-wise losses. The framework preserves the basic representation format of hybrid NVRs and can therefore be applied to different backbones. Experiments on the UVG
dataset at resolutions of \(1080\times1920\) and \(810\times1440\) show that the proposed approach improves reconstruction quality over the corresponding NVR baselines.
Read the full abstract ▾
Robotics
PINGU: Extending Air-Bearing Spacecraft Emulators with Open-Source Actuators and Learned Control for Contact-Rich Proximity Operations
Ricard Marsal I Castan, Akiyoshi Uchida, Aman Arora et al. (+7 more)
Low-cost planar air-bearing testbeds have matured into a standard proxy for free-flying spacecraft GNC, but they remain largely thruster-only and are rarely equipped for contact-rich, inertia-coupled manipulation. Building on the open-source ATMOS testbed, we contribute a reaction wheel and two force/torque-sensed robotic arms (LEVION) with interchangeable end-effectors, integrated as first-class control actuators through a unified ROS 2 abstraction layer. On top of the software stack we build a reinforcement-learning training environment and digital twin, and a controller that exploits these added degrees of freedom, letting classical optimal controllers and learned policies be swapped on the same hardware without modification. We validate the integrated system, PINGU, across four
benchmark tasks: point-to-pose
navigation (classical LQR vs. sim-to-real PPO), dynamic disturbance rejection under arm-induced center-of-mass shifts, reaction-wheel momentum stabilization, and force-controlled docking. The results show that these additions extend an ATMOS-class emulator into the contact-rich regime and bridge classical optimal control and reinforcement learning on one reproducible platform.
Read the full abstract ▾
All authors Ricard Marsal I Castan, Akiyoshi Uchida, Aman Arora, Pedro Lima, Matteo El-Hariry, Anrej Orsula, Francesco Grella, Antoine Richard, Cedric Pradalier, Miguel A. Olivarez-Mendez
Computer Vision
Modeling Clinical Workflow for SYNTAX Scoring from Coronary Angiography Videos
Suzhong Fu, Jingqi Dong, Xuan Ding et al. (+4 more)
The SYNTAX score is a clinically established tool for assessing anatomical lesion complexity in coronary artery disease and guiding subsequent treatment. However, automated SYNTAX scoring is commonly formulated as a direct regression problem from coronary angiography
videos to patient-level scores. In this work, we reformulate SYNTAX scoring as a vessel segment identity-preserving anatomical
reasoning problem and propose a hierarchical modeling framework that explicitly aligns learning with the clinical workflow. Our approach maintains vessel segment identity across frames and views, estimates stenosis severity at the segment level, and aggregates evidence hierarchically according to coronary anatomy. Simultaneously, to address the scarcity of domain-specific data, we integrate and complete multiple public coronary angiography
datasets, constructing a large-scale resource featuring completed vessel
segmentation and derived structural annotations. Experiments demonstrate that vessel segment-level stenosis
embedding enhances explanatory power and reduces prediction variability compared to baseline models, with the R^2 score improving by 0.201 and dev STD decreasing by 18.4%. These results highlight the necessity of structure-aligned modeling for reliable and stable automated SYNTAX scoring from multi-view coronary angiography
videos. The GitHub link is https://github.com/VersaceSu7/SYNTAX_score_777.
Read the full abstract ▾
All authors Suzhong Fu, Jingqi Dong, Xuan Ding, Rui Sun, Yiming Yang, Shuguang Cui, Zhen Li
High Energy Physics (Phenomenology)
Investigation of the difference in the angular distributions of $Z\to \ell^{+}\ell^{-}$ events produced in quark-antiquark, quark-gluon and gluon-gluon collisions
Arie Bodek, Giulia-Maria Bulugean, Aran Garcia-Bellido et al. (+3 more)
Measurements of the angular distributions of Z
boson-decay
leptons at the Large
Hadron Collider (LHC) are of interest in QCD studies and precision electroweak measurements. Z
boson production at the LHC is dominated by
quark-
gluon processes and the
quark-antiquark process accounts for only 40% of the cross section. We investigate the theoretical predictions (using the POWHEG-minnlops event generator) for the difference in the angular distributions of $ pp \to \gamma^{*}/Z \to \ell^+\ell^-$ ($ \mu^+\mu^-$ or $e^{+}e^{-}$) events produced via
quark-antiquark,
quark-
gluon , and
gluon-
gluon processes. The study is done for
proton-
proton collisions at the Large
Hadron Collider (LHC) at $\sqrt{s} = 13$ TeV. We investigate the angular coefficients $A_0$ and $A_2$ (for $\theta_{CS}$ and $\phi_{CS}$ distributions in the Collins-Soper frame) for the different processes as a function of the Z
boson rapidity, transverse momentum and final state jet multiplicity, and propose how these can be investigated experimentally at the LHC.
Read the full abstract ▾
All authors Arie Bodek, Giulia-Maria Bulugean, Aran Garcia-Bellido, Hyon San Seo, Rhys Taus, Un-Ki Yang
Computation and Language
Paragraph Boundaries Are Not White Space:Compression Depth as the Signature of Hierarchical Structure
Shuyang Xiang
Standard positional encodings treat position as a one-dimensional reading-order coordinate, but reading order alone does not determine hierarchical textual structure. We use a hierarchical rotary positional encoding (hRoPE) that represents paragraph, sentence, and
token indices as separate channels, hold the
token sequence fixed, intervene on the paragraph coordinate $p_1$, and measure cross-paragraph
attention with a
token-distance-exact
estimator.
Attention is compressed in every corpus, but compression alone is not diagnostic of true structure: an architecturally identical channel with density-matched random labels is compressed too. What distinguishes real structure is the depth of compression: the real-versus-random depth gap is resolvable in two of three corpora and not in the third, and real-structure depth varies more strongly across corpora than the control's. Comparing eight corpus-only quantities across three constructs (lexical persistence, paragraph length,
embedding-based coherence), none fully reproduces the cross-corpus ordering of depth, though
embedding-based coherence comes closest. At the paragraph level that relation transfers as a common slope, but with the opposite sign to the corpus-level ranking: controlling for length, diversity, and position, paragraphs with more similar neighbors compress less deeply, with no detectable slope difference between any pair of corpora, while length, lexical-diversity, and position effects remain corpus-specific. Compression depth, not its location, is the reproducible signature of genuine paragraph structure in our setting.
Read the full abstract ▾
Sound
ArtifactBench: Lineage-Aware Evaluation of AI-Generated Music Detectors under Distribution Shift
Heewon Oh
AI-generated music detectors are commonly compared using aggregate scores on
benchmarks whose training overlap, generator lineage, source provenance, and audio-transformation history are only partially observable. This paper introduces ArtifactBench, a lineage-aware evaluation suite for measuring detector behavior across generator families and versions, real-music domains, collection-cohort shift, and inference coverage. The
benchmark groups source recordings and their derived variants by content identity, separates calibration from final testing, records inference failures independently from classification errors, and reports source-level performance with uncertainty in addition to aggregate metrics. We evaluate multiple publicly available detectors under a version-pinned common protocol and examine how leakage control, cohort availability, threshold policy, and model-specific missingness alter measured performance and model ranking. On the 562-track common-success test intersection, ArtifactNet obtains 0.982 AUROC and 0.918 balanced accuracy, compared with 0.761/0.776 for the public Deezer detector; SpecTTTra and CLAM fall below 0.30 AUROC under this shifted cohort. These results also expose substantial generator- and real-domain shifts that aggregate scores alone conceal.
Read the full abstract ▾
Machine Learning
Physics-residual machine learning predicts oxygen-evolution catalyst activity beyond the training range from sparse polarization measurements
Yong-Woon Kim, Jihyeok Lee, Sungtae Park et al. (+2 more)
Discovery campaigns for oxygen
evolution reaction catalysts repeatedly choose, make and measure catalysts. High-throughput platforms stop polarization curves below potentials that damage the catalyst, so the endpoint, the activity at a target potential or current density, often lies beyond the measured window, and the catalysts of most interest are more active than any measured before. Existing methods do not predict these endpoints accurately when few or no endpoints of a new library have been measured. Here we present physics-residual machine learning (PR-ML), which predicts each endpoint as the sum of a Tafel term, computed from the catalyst's own measured curve with an estimated slope, and a residual term learned from labelled catalysts. In twelve Ni-Pd-Pt-Ru thin-film libraries, the current density at 1.70 V$_{RHE}$ was predicted from the currents at 1.40 and 1.55 V$_{RHE}$. Fitted only on earlier libraries, with ridge regression as the residual learner, PR-ML predicted the Ni–Ru library, whose currents mostly exceed theirs, with a mean absolute error of 0.194 mA cm$^{-2}$, against 1.330-1.882 for data-driven models. With five endpoints from the new library and extremely randomized trees as the residual learner, PR-ML gave a similar error, which the same learner used alone reached only with 20, and identified 63-83% of the catalysts more active than the best labelled catalyst, against 2%. In two independent
datasets, this fraction rose from at most 1% to 33-95%. Our approach supplies catalyst selection with accurate endpoints beyond the measured part of each curve and above all earlier measurements.
Read the full abstract ▾
All authors Yong-Woon Kim, Jihyeok Lee, Sungtae Park, Sooseok Choi, Yung-Cheol Byun
Computer Vision
SewFusion: Tailored Generation of Topology and Panel-Level Geometry for Sewing Patterns
Jiaxin Lin, Xiao Pan, Hangjie Yuan et al. (+3 more)
Generating sewing patterns from images and text requires modeling a heterogeneous representation composed of discrete
topology and continuous geometry. Existing methods mainly follow two paradigms:
diffusion-based methods enable holistic geometry generation by converting the entire pattern into a continuous representation, but weaken discrete
topology modeling; in contrast, autoregressive methods preserve discrete
topology through next-
token prediction, but tie continuous geometry regression to
token-level hidden states with limited panel-level context. To bridge this gap, we propose SewFusion, a unified autoregressive framework that adopts tailored generation mechanisms for discrete
topology and panel-level continuous geometry, using next-
token prediction for the former and flow matching for the latter. To support panel-level continuous geometry generation, we introduce a Panel Geometry VAE that learns a fixed-size latent space for variable-length panel geometry, together with Panel Geometry Flow for latent generation. We further propose Panel-Forcing to reduce the training–inference mismatch in
topology context and improve robustness to
topology prediction errors. Extensive experiments on SewFactory and GCD-MM demonstrate that SewFusion consistently outperforms previous state-of-the-art methods across various settings, achieving +6.36% Panel Accuracy, +11.30% Stitch Accuracy, and -1.90 Vertex L2 error in the image-text-based generation setting.
Read the full abstract ▾
All authors Jiaxin Lin, Xiao Pan, Hangjie Yuan, Luyan Liang, Wan Li, Daquan Feng
Machine Learning
Preserving Geometric Integrity in Graph Prompting via Measure-Constrained Optimal Transport
Xiangyu Wang, Shuo Wang, Ruiyi Fang et al. (+1 more)
Graph
prompt learning enables parameter-efficient adaptation of frozen Graph Neural Networks to downstream tasks through lightweight
prompt parameters. As routing becomes increasingly node-adaptive, however, independently optimized local decisions can collectively concentrate assignment mass on a small subset of a finite shared
prompt bank, even when individual node–
prompt matches remain locally meaningful. We propose MINT (Measure-INtegrity Transport), an entropically regularized optimal transport framework that formulates node-to-
prompt adaptation as a globally coupled allocation problem. The transport cost favors local
geometric compatibility, while a prescribed
prompt-side marginal explicitly controls graph-wide
prompt utilization. We further derive an exact variance decomposition that separates
prompt-side
geometric variance into retained
prompt-update variation and within-node barycentric dispersion, together with a conditional stability bound for the frozen-encoder forward map. Across standard citation networks and additional heterophilic graphs, MINT remains competitive in few-shot adaptation. Controlled and end-to-end experiments further distinguish the roles of routing and
topology: fixed-marginal routing controls graph-wide
prompt utilization and has measurable end-to-end effects on citation networks, while
topology augmentation provides a complementary, graph-dependent mechanism for addressing structural mismatch. Code is available at https://github.com/Ga1axy0051/MINT.
Read the full abstract ▾
All authors Xiangyu Wang, Shuo Wang, Ruiyi Fang, Zhao Kang
Numerical Analysis
A novel monotonous finite volume element scheme for convection dominant diffusion problem
Cunyun Nie, Xiaoling Chen, Zhujun wang et al. (+2 more)
One novel monotonous finite volume element (MFVE) scheme is put forward for convection dominant
diffusion problem. The main contributions of this paper include four aspects. Firstly, one upwind sub-control volume, called upwind volume, is introduced for the discretization of the convection term similar to the upwind element, which leads to the upwind property. Secondly, one first-order linear discrete operator for the convection term in the balance equation is directly obtained by numerical integration in upwind volume, which is different from the classical finite volume methods.Thirdly, one asymptotic expansion of gradient functions in the dual element of each node is derived by one extension interpolation function. Then, one second-order nonlinear discrete operator for the convection term is constructed by the asymptotic expansion and error estimation of its FVE solution, where some perturbed coefficients are skillfully put into the expansion, which serves as a high-order correction. Fourthly, one novel two-order MFVE scheme, also accompanied by another one-order MFVE, is designed for convection-dominated
diffusion problems together with one positivity-preserving finite volume element (PFVE) discrete diffusive operator. The L2 norm of the error of the approximate solution is derived. Finally, numerical results validate theoretical conclusions.
Read the full abstract ▾
All authors Cunyun Nie, Xiaoling Chen, Zhujun wang, Zhikun Tian, Chengjie Xia