Peer Revenue Publish or profit

Papers

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

All of arXiv

Saturday 5 September

Computation and Language

STQA: A Benchmark for Stock-Focused Tabular Question Answering over Historical and Forecasted Data

Stock market analysis inherently requires composite reasoning over historical records and future projections, yet existing benchmarks remain fragmented across isolated tasks. We introduce STQA (Stock-focused Tabular Question Answering), an end-to-end benchmark designed to systematically evaluate natural-language question answering over historical data, numerical forecasts, and forecast-based reasoning. Built on a large-scale financial dataset, STQA covers 4,417 stocks and contains 31,400 question-answer pairs derived from expert-crafted templates, accompanied by fine-grained intent and slot annotations. To operationalize this benchmark, we present SQFRS (Stock Query-Forecast-Reasoning System), an agent-based unified framework that orchestrates SQL retrieval and time-series forecasting tools. Experiments demonstrate that while current large language models perform well on historical queries, forecast-based reasoning poses a substantial challenge, revealing critical bottlenecks in tool coordination and reasoning under uncertainty. The dataset and code are available at https://github.com/xuxubaobaoan/STQA_Project. STQA thus serves as a rigorous testbed for future research on trustworthy, tool-augmented financial agents.
Read the full abstract

All authors Baoxu An, Wenmian Yang, Zhensheng Wang, Weijia Jia

Human-Computer Interaction

MM-SVGEdit: A Multimodal-Driven SVG Editing for UI Design

In the field of UI design, Scalable Vector Graphics (SVG) is widely used as a design medium. However, traditional SVG editing techniques have high entry barriers and require cumbersome manual iteration, while LLM-based editing solutions suffer from low accuracy and poor user controllability. To address these issues, we propose MM-SVGEdit, a multimodal-driven SVG editing approach that integrates traditional SVG editing and LLM-based methods. We introduce a two-stage strategy in which visual grounding is followed by modification. Both stages support two interaction modalities: natural language instructions and direct manipulation (mouse and keyboard). We trained and evaluated MM-SVGEdit on a self-constructed dataset of 14,476 question-answer pairs generated from UIs, covering 11 types of editing operations on both single and multiple UI targets. The results show that MM-SVGEdit improves SVG editing accuracy, efficiency, and user-perceived control while reducing token consumption and response time.
Read the full abstract

Nuclear Theory

Quantum Theory Angular Momentum; Electronic version 2026

This is an electronic version of the open source book "Quantum Theory of Angular Momentum" by D. A. Varshalovich, A. N. Moskalev, V. K. Khersonskii, original copyright o̧pyright} 1988 by World Scientific Publishing Co. Pte. Ltd, with the CC-BY 4.0 open source ebook funded by SCOAP. We have made appropriate modifications for the modern age–such as python-based symbolic manipulators to replace algebraic and numeric the tables. The new version contains a number of small modification, even though it is close to the original, with additional indices, hyperlinking, redrawing of figures. There has been a substantial check of the results presented, supported by the use of AI, even though all mistakes are the author's responsibility. All material is available from the book's github site (https://github.com/nwalet/Varshalovich/), where new updated versions will also be presented.
Read the full abstract

Robotics

Where Success Breaks: Failure-Boundary Learning for Robust Vision-Language-Action Models

Vision-language-action (VLA) models adapted through supervised fine-tuning (SFT) inherit a structural asymmetry: expert demonstrations teach the policy where success behavior lies, but provide no signal about where it ceases to be reliable. We argue that robust VLA adaptation should therefore be viewed not as further demonstration fitting, but as Failure-Boundary Learning – the problem of Discovering, Localizing, and Shaping the boundary between recoverable deviations and task failure. To instantiate this view, we propose DLS: built on a real-grounded behavioral prior from few real demonstrations and simulated co-training, DLS discovers failure boundaries at scale through on-policy digital twin rollouts. Rather than reducing each rollout to a binary label, semantic progress localization uses privileged simulator states to assign progress-aware signals that capture where the failure boundary is crossed, not merely whether. These signals drive directional boundary shaping in the flow dynamics – reinforcing success-producing denoising directions and suppressing failure-producing ones, without action likelihoods or auxiliary critics. Across real-robot manipulation tasks, DLS improves robustness over SFT and online RL baselines, especially under randomized initial states and unseen visual conditions.
Read the full abstract

Analysis of PDEs

Qualitative analysis of positive radial singular solutions on hyperbolic space

We investigate positive radial solutions with an isolated nonremovable singularity for the semilinear elliptic equation \begin{align*} \Delta_{\mathbb{H}^N} u+\lambda u+u^p=0 \qquad\text{in }\mathbb{H}^N\setminus\{Q\}, \end{align*} where $N\geq 3$, $p>1$, $\lambda\le \frac{(N-1)^2}{4}$, and $Q\in\mathbb{H}^N$ is a prescribed pole. Our purpose is to describe how the locally Euclidean singular behavior near ${\{Q}\}$ interacts with the genuinely hyperbolic dynamics at infinity, and how this interaction changes across the Serrin and Sobolev critical exponents. For $\frac{N}{N-2}\le p<\frac{N+2}{N-2}$, we construct a family of positive radial singular solutions selecting the fast exponential mode at infinity. At the pole, these solutions exhibit the logarithmically corrected fundamental-solution profile when $p=\frac{N}{N-2}$, and the standard power-law profile when $\frac{N}{N-2}<p<\frac{N+2}{N-2}.$ At the conformally invariant pair $(p,\lambda)=(\frac{N+2}{N-2},\frac{N(N-2)}{4})$, we obtain an explicit singular solution and nonconstant Fowler-type solutions, and we classify all positive radial solutions with a nonremovable singularity via the positive periodic orbits of an associated autonomous equation. In the supercritical regime $p>\frac{N+2}{N-2}$ with $\lambda\le\frac{ N(N-2)}{4}$, we prove existence and uniqueness of the global positive radial singular solution, derive its two-term local asymptotic expansion near the pole, and establish a sharp trichotomy for its behavior at infinity.
Read the full abstract

Earth and Planetary Astrophysics

Shaving the Outskirts of Planetary Systems Probed by Roman via Stellar Flybys

The recently launched Nancy Grace Roman Space Telescope (Roman) will soon begin observations that promise to revolutionize our understanding of exoplanet demographics. Roman's Galactic Bulge Time-Domain Survey (GBTDS) will observe the Galactic Bulge via transit and microlensing techniques, providing a unique opportunity to study hot Jupiters as well as long-period/free-floating exoplanets in a dense and kinematically hot stellar environment. In this Letter, we quantify how repeated, impulsive stellar flybys truncate the outskirts of planetary systems in the Bulge. We find that flybys result in ejections on $\sim$Gyr timescales for wide orbits (${\gtrsim} 250\ \text{au}$) around low-mass hosts. Provided these wide orbits are populated, these ejections likely contribute to the population of free-floating planets probed by Roman. Surviving planets outside of ${\sim} 10$ au are placed on eccentric and misaligned orbits. We show that, though systems suffer many close encounters, the evolution is dominated by the single strongest perturbation; planetary orbits (even those that evade ejection) undergo superdiffusion (i.e. a Lévy flight). We comment on the implications of our results for the origin of hot Jupiters amenable to $\textit{Roman}$'s transit search. Our work serves as a quantitative baseline from which future dynamical studies can build to better understand the evolution of planetary systems in the Galactic Bulge.
Read the full abstract

Quantum Physics

Measurement-Based Feedback of Open Quantum Systems: A Control-Theoretic Review and Tutorial

This review develops a control-theoretic perspective on measurement-based feedback for continuously monitored open quantum systems, with the main analysis focused on finite-dimensional systems governed by diffusive stochastic master equations. We introduce the relevant state-space, invariant-subspace, and quantum non-demolition structures, interpret quantum filtering as nonlinear observer dynamics, and review open-loop asymptotics, filter stability, state-feedback stabilization, robustness, and reduced-order observer-based control. Particular emphasis is placed on a recurrence–contraction framework, in which Hamiltonian feedback removes non-target invariant obstructions while measurement-induced dynamics provide local exponential contraction. Although the detailed analysis is developed for finite-dimensional diffusive models, the underlying measurement–estimation–feedback architecture is relevant across a broad range of quantum platforms. We further discuss implementation challenges and open problems involving scalable estimation, sampling and delay, adaptation, hybrid and non-Markovian dynamics, practical stability, optimal control, and learning-based design. By organizing these developments around invariance, estimation, recurrence, and contraction, the review provides a tutorial bridge between measurement-based quantum feedback and nonlinear stochastic control.
Read the full abstract

All authors Weichao Liang, Jing Zhang, Gaoyue Guo, Daoyi Dong

Group Theory

Property $P_{\text{naive}}$ for non-orientable big Mapping Class Groups

We give a criterion for property $P_{\text{naive}}$ for subgroups of orientable big mapping class groups and apply it to mapping class groups of non-orientable infinite type surfaces through the orientation double cover.

Machine Learning

Sparse Incident-Cluster Learning for 12-hour Port Flood Pre-warning in Digital-Twin Analytics

Port flood digital twins require analytics that warn operators before disruption, but official warning incidents are often few and adjacent observations are temporally dependent. Row-level classification can therefore overstate performance by placing windows from the same event in both model-development and evaluation data. We formulate 12-hour port flood pre-warning as an incident-cluster learning problem and evaluate a digital-twin analytics module using eight-point water-level histories, prediction-time contextual covariates, and interpretable short-window dynamics. The protocol combines fold-specific sparse feature selection, warning-cluster grouping, negative-label controls, 100-repeat random top-k controls, and alert-episode evaluation. Liverpool is the primary four-cluster case study, with harmonised Humber/Hull-proxy and Wessex South data used for protocol-transfer checks. Across the Liverpool folds, the top-10 ElasticNet model achieves mean F2 = 0.696, compared with 0.633 without top-k truncation and 0.681 for full-feature weighted XGBoost. It is the strongest ElasticNet variant, remains competitive with the nonlinear reference using only ten predictors, and exceeds the repeat-level 95th percentile of broad and same-family random subsets. Contextual covariates provide a strong prediction-time anchor, complemented by physically interpretable local dynamics. Historical replay converts risk scores into alert episodes and measures alert duration and false-episode burden. The result is an offline-evaluated analytics and validation module designed for integration into a port digital twin.
Read the full abstract

All authors Jie Zhang, Qiang Ni, David Windridge, Huan X. Nguyen

Computation and Language

From Two Passes to One: Compact and Efficient Target-Stance Extraction

Target-Stance Extraction (TSE) is the task of predicting both the target (or topic) of an author's writing and the author's stance toward it. Existing approaches to TSE use a sequential pipeline of two separate neural models: one to identify the target and another to determine the stance. We present a one-pass, joint architecture that predicts both in a single forward pass, reducing trainable parameters by nearly 50% with only a 4-7 F1 point tradeoff in performance. We further demonstrate that standard target-scrubbing practices artificially suppress target prediction accuracy. Retaining explicit target mentions, as in real-world deployments, improves F1 by at least 6 points across both target classification and target generation settings. These improvements allow for significantly easier integration of TSE in downstream applications such as public opinion tracking.
Read the full abstract

Machine Learning

DataFlex-RL: An Evaluation Platform for RLVR Data Policies

Data policies for reinforcement learning with verifiable rewards (RLVR) determine which rollouts are used, how strongly they are weighted, and which domains contribute to subsequent training batches. We introduce DataFlex-RL, an evaluation platform for comparing these choices under a common GRPO recipe. Our primary experiment evaluates 13 configurations across 12 matched seeds using Qwen2.5-7B-Base and 12 mathematics, logic, and science benchmarks. Uniform GRPO improves the domain-balanced average accuracy by 7.76 percentage points over the untrained checkpoint. None of the eight rollout-selection or reweighting methods achieves a paired 95% confidence interval that excludes zero relative to uniform sampling, and none of the three adaptive mixtures outperforms a fixed equal mixture at the same level of precision. A corrected 12-seed extension on Llama-3.1-8B-Base places the additional methods on the same score scale as the original controls, but does not reveal a consistent winner in terms of observed mean performance. We also quantify evaluation sensitivity by rescoring nine Qwen2.5-7B-Instruct runs using a math-heavy six-benchmark summary, consisting of five mathematics benchmarks and GPQA-Diamond but no logic benchmark, and comparing it with the domain-balanced 12-benchmark summary. The resulting rankings are negatively correlated, with a correlation coefficient of -0.33, whereas summaries that retain all 12 benchmarks largely agree. Across the controlled settings studied here, changing the data policy measurably changes the training process but does not produce a reproducible improvement over uniform training.
Read the full abstract

All authors Hao Liang, Mingrui Chen, Hengyi Feng, Meiyi Qiang, Wentao Zhang

Machine Learning

FANS: Federated Adaptive Network Search Learning for Heterogeneous Devices

Heterogeneous Federated Learning (HFL) aims to train models across devices with diverse resource budgets while preserving data privacy. Existing HFL methods typically bind training to a small predefined menu of model configurations, which limits architectural coverage. To address this bottleneck, we introduce Federated Adaptive Network Search (FANS), a hypernetwork-based framework that learns a shared architecture space rather than a fixed set of client models. To optimize this shared space efficiently, we propose the Federated Parallel Scaling (FPS) algorithm, which jointly trains multiple sampled subnetworks in parallel with self-distillation so that larger sampled subnetworks can supervise smaller ones during local updates. We evaluate FANS on CIFAR-10, CIFAR-100, and MNLI using ResNet-18, DenseNet-121, and BERT-base, respectively. Across all benchmarks, FANS expands the feasible subnetwork pool by orders of magnitude (e.g., 4,680 candidates for ResNet-18 vs. 4 in existing methods) and improves the average accuracy-efficiency trade-off relative to representative HFL baselines. Device heterogeneity is emulated through resource tiers, and evaluation covers accuracy, parameter count, and MACs.
Read the full abstract

All authors Jiaxin Zhang, Xingwei Wang, Bo Yi, Liang Zhao, Alireza Furutanpey, Ziyi Chen, Qiang He, Keqin Li, Schahram Dustdar

Instrumentation and Detectors

Simulation Study of the Design and Spatial Resolution of a Novel AC-coupled Cross Strip LGAD

AC-coupled Low Gain Avalanche Diode (AC-LGAD), evolved from the standard LGAD technology, are silicon detectors characterized by exceptional temporal and spatial resolutions. This work proposes a novel sensor named AC-coupled Cross Strip LGAD (CS-LGAD) structure with dual-layered overlapping strip electrodes. It can achieve two-dimensional particle position measurement with low readout density while also inheriting the high temporal resolution of the LGAD. The sensor structure, doping and current-voltage (I-V) characteristics of the sensor were investigated through TCAD simulations. Additionally, Monte Carlo simulations were employed to simulate the signal response of CS-LGAD to minimum ionizing particles (MIPs). The response of MIPs hitting different positions was simulated, and the reconstruction of particle coordinates (x and y) was achieved based on the sharing of charges on the dual-layer electrodes. The spatial resolution might reach 3.6% and 5.5% of the pitch size in the x and y directions, respectively, according to the simulation. The proposed CS-LGAD is a promising 4D detector with low mass and low readout density.
Read the full abstract

All authors Jingkang Xu, Mengzhao Li, Zhiliang Hu, Tianya Wu, Mei Zhao, Zhijun Liang, Lihua Mo, Tianjiao Liang

Quantum Physics

Battery Locality Is Necessary in Noncommuting Quantum Charging Bounds

Bounds on the charging power of quantum batteries in direct charging protocols are often interpreted as charging-side constraints. Although the locality of the charging Hamiltonian is known to restrict attainable power in certain protocols, whether battery locality is itself operationally necessary has remained unresolved. Here we show that it is necessary for spin batteries whose interactions do not pairwise commute. We construct commuting and noncommuting high-locality batteries with identical interaction supports and local energy scale, driven by the same charging. The commuting battery saturates the battery-locality-independent bound, whereas the noncommuting battery exceeds it by a parametrically growing factor, demonstrating that this bound cannot apply generally. The enhanced norm of the commutator is dynamically attained from a suitable joint state supported entirely in the battery ground-energy sector, although this state may be correlated with the auxiliary fermionic degrees of freedom. Our results establish battery locality as an operational resource enabled by noncommuting battery interactions.
Read the full abstract

Cryptography and Security

Beyond the Prank: The Hidden Expertise of TSS Scambaiters

This research studies how Technical Support Scams (TSS) are being countered by a uniquely dedicated community of volunteer counter-fraud operatives. Using a careful subject selection strategy, we interviewed 17 individuals who actively engage in TSS scambaiting activities in order to obtain insight into their motivations, the operational methods of the scammers they combat, the undocumented nuances of effective scambaiting action, and the various challenges scambaiters face. In our analysis, we find a community rich not only with insight into offenders, but with technical and operational expertise that is often lacking in research efforts targeting these same populations. At the same time, we find key areas where the community could be better supported and enabled. We discuss the implications of our findings for both future research and community protection strategies.
Read the full abstract

All authors Saleh Alsyefi, Anish Chand, Matthew Edwards, Phani Vadrevu

Neural and Evolutionary Computing

Programmable Cellular Automata

Cellular automata is a local computation paradigm where complex behavior can arise from local interactions between simple functions. This paradigm has been used to explain many systems such as biological processes, traffic simulation, computer networks, etc. In games, cellular automata have been used in games such as SimCity and for the generation of spatial content such as caves or dungeons. However, creating effective local rules is hard and unintuitive. Cellular automata can be effectively evolved, but may still be hard to interpret. In this work, we introduce the concept of programmable cellular automata, where we represent the system as Python code. We also modularize the cellular automata into local functions and a decision function. Local functions take a local neighborhood and return a value, while the decision function takes the output of the local functions and decides the value of the next state. Separating the cellular automata into modules written in Python helps with understanding how these systems are working. We also explore adding global functions where they take the whole state and compute a function from it. We tested generating levels for three different games from the PCG Benchmark. The results showed that global functions decrease the number of iterations that cellular automata need to solve a problem, and that we cannot find solutions for some problems with purely local functions. Looking into the generated functions, we can see common functions that have been used in different experiments, which not only helps us understand the generator but also helps us understand these games better and what is important for them.
Read the full abstract

All authors Ahmed Khalifa, Muhammad Umair Nasir, Matthew Siper, Steve James, Julian Togelius

Audio and Speech Processing

SETEAB: Multiscale approach with Squeeze-and-Excitation Temporal Enhanced Aware Block for Speech Emotion Recognition

This paper proposes a novel lightweight multiscale architecture for speech emotion recognition (SER) with three key innovations. First, a depthwise convolution-based subsampling module is introduced to reduce model size and computation while preserving salient emotional cues. Second, a Squeeze-and-Excitation block is integrated to enhance channel-wise recalibration and improve representation robustness. Third, a new Temporal Enhanced Aware Block is designed to strengthen temporal dependency modeling and produce more discriminative emotion-aware features. The proposed model is explicitly designed to jointly improve compactness, recognition performance, and generalizability. Experiments on benchmark SER datasets show that our method achieves higher accuracy with reduced computational complexity, while also delivering stronger cross-corpus performance than most recent advanced networks for SER.
Read the full abstract

Machine Learning

VERPO: Verified Evidence Regularized Policy Optimization

Verifiable rewards improve language models through reliable task-level feedback, but methods based on Group Relative Policy Optimization (GRPO) apply a sequence-level advantage uniformly across all tokens. This coarse credit assignment reinforces or penalizes entire responses without identifying which local decisions to preserve, reinforce, or revise. Conversely, evidence-conditioned self-distillation provides denser token-level supervision, yet teacher imitation can transfer stylistic artifacts and miscalibrated confidence that destabilize training when misaligned with task success. We introduce VERPO, which converts evidence-conditioned guidance into reward-aligned token-level credit assignment while retaining the outcome objective. VERPO decomposes teacher guidance into an evidence-free reference term and signed, evidence-induced corrections at each token. A stopped controller combines selective acceptance, token-wise localization, and cost-aware scaling by balancing alignment with the local GRPO update direction against Fisher movement cost. Furthermore, we introduce Fisher Evidence Contrast (FEC), which attenuates nuisance shifts along an estimated evidence-presence direction through a regularized projection. Across five scientific reasoning and tool-use tasks, VERPO prevents optimization collapse and consistently achieves the highest multi-task average across model backbones, yielding marked improvements particularly on smaller models over strong baselines. Qualitative diagnostics confirm that token acceptance selectively targets reasoning bottlenecks consistent with local reward alignment and Fisher movement cost.
Read the full abstract

All authors Haijiang Li, Chengyu Lv, Yi Zhang, Rui Qian, Zhibing Zhang, Xiangqing Shen, Junjie Yang, Yuchen Zhang, Wenyuan Jiang, Hanqing Hu, Cangqi Zhou

Computer Vision

PASTEL: Panoramic Alignment for Monocular 4D Scene Reconstruction

Reconstructing 4D scenes from casually captured monocular video is vital for applications in virtual reality (VR) and embodied AI. Recent advances in 4D reconstruction and novel view synthesis have substantially propelled this capability. However, existing reconstruction methods generally cannot recover regions beyond visible camera limits. Consequently, we introduce a new paradigm that achieves 4D scene synthesis by combining visible-region reconstruction from monocular input with invisible-region generation beyond observable camera boundaries. We present Panoramic Alignment for Strategic Exploitation of Generative Priors (PASTEL). Specifically, PASTEL proposes panoramic scene alignment, a novel representation that reformulates the intractable 3D "invisible region" exploration into a tractable 2D directional trajectory planning. This is achieved by reducing the viewpoint planning from 6-DoF search to a 2D directional search with explicit visibility boundaries. By operating within this panoramic space, our method strategically identifies camera trajectories that maximize exploration beyond observable boundaries while minimizing viewpoint deviation. Experimental results show that PASTEL can not only extrapolate plausible scene content beyond the observable boundaries of input monocular videos, but also substantially boost monocular 4D reconstruction performance. PASTEL outperforms the previous state-of-the-art method by 0.9dB in full-image PSNR on the DyCheck IPhone dataset.
Read the full abstract

All authors Yuankun Yang, Yi Wei, Bo Bai, Wenyang Zhou, Li Zhang

Statistical Methodology

Causal DAG Identification for Count Data via Poisson Thinning Structural Equation Models

Count-valued variables arise in many scientific and applied settings, yet explicit structural models that allow full identification of causal DAGs from observational data remain limited. The Poisson branching structural causal model (PB-SCM) provides a count-valued analogue of linear structural equation models using binomial thinning and independent Poisson exogenous variables, but its causal DAG is generally only partially identifiable. Building on this framework, we propose the Poisson thinning structural equation model (PT-SEM), which replaces binomial thinning in PB-SCM with Poisson thinning and allows node-wise exogenous distributions from diverse count-distribution families. Under node-wise regularity conditions, we establish identifiability of the causal DAG, the thinning coefficients, and the node-wise exogenous distributions. The same identification analysis extends to binomial thinning, yielding full identifiability whenever every nonsink has non-Poisson exogenous noise. We further develop a structure learning algorithm that optimizes, via dynamic programming, a BIC score based on local likelihoods evaluated at plug-in moment estimates, and establish its consistency for DAG selection. Simulations demonstrate favorable performance in DAG recovery and thinning-coefficient estimation, and a real-data application illustrates the practical utility of PT-SEM.
Read the full abstract

General Relativity and Quantum Cosmology

Regular black holes from dynamical-tension string hedgehogs

A spherical cloud of radial strings with non-zero constant tension has an energy density proportional to $r^{-2}$ and therefore produces a singular conical geometry rather than an asymptotically flat regular black hole. We show that this obstruction can be removed within the modified-measure formulation of strings. A continuum of radial worldsheets whose tensions are generated by a composite bulk tension scalar reduces, on a constrained $\mathrm{SO}(3)$ hedgehog branch, to Einstein gravity coupled to a nonlinear function $K(Y)$ and an auxiliary three-form. This construction gives model-independent conditions directly in terms of the radial tension $\mathcal{T}(r)$. Finite curvature at the centre requires $\mathcal{T}=\mathcal{O}(r^2)$, while the null energy condition forbids a faster leading power. Thus a positive regular centre satisfying the null energy condition is necessarily de Sitter. Finite ADM mass requires the tension to be screened at infinity, whereas the dominant energy condition cannot hold globally for a non-trivial positive profile screened faster than $r^{-2}$. An exact cubic profile realises these properties and gives a continuous family of asymptotically flat geometrically regular black holes whose first correction to Schwarzschild appears at order $r^{-4}$.
Read the full abstract

Combinatorics

On the real-rootedness of Ehrhart $h^*$-polynomials of $(132,213)$-avoiding permutahedra

Let $P_d$ be the convex hull of the permutations of $\{1,\ldots,d\}$ that avoid the patterns $132$ and $213$, and let $H_d(t)$ be its Ehrhart $h^*$-polynomial. We prove that all zeros of $H_d(t)$ are real and negative for $3\le d\le1000$ and for every $d\ge2^{62}$; for $d\le2$ the polynomial is constant. Expanding $H_d(t)$ in refined Eulerian polynomials does not suffice, since a coordinate is already negative at $d=7$. Let $A_{d,i}(t)$ enumerate all permutations of $\{1,\ldots,d\}$ with first entry $i$ by number of descents. For $c\ge0$, we prove that every nonnegative linear combination of $A_{d,1},\ldots,A_{d,d}$ and the adjacent differences $A_{d,r+1}-cA_{d,r}$, $2\le r\le d-2$, is real-rooted for all $d\ge4$ simultaneously if and only if $c\le2-\sqrt3$. This reduces real-rootedness to the positivity of explicit weighted tails of the coordinates. An exact certificate checks these inequalities for $d\le1000$, and coefficient estimates based on Darroch's theorem on the mode of a sum of Bernoulli variables prove them for $d\ge2^{62}$; the range $1001\le d<2^{62}$ remains open. Along the way we identify $P_d$, up to lattice equivalence, with twice a chain-poset permutahedron and with the Newton polytope of a discriminant, which transfers the known cube structure and volume of these models to $P_d$. We also show that the coefficient of $t^{d-1}$ in $H_d(t)$ counts strong tournament score sequences for $d\ge3$, confirming a conjecture of Rehberg in this range.
Read the full abstract

Human-Computer Interaction

Flawed but Memorable: Student Critical Reception of Interest-Personalized GenAI Analogies in Computing Education

Motivation: Undergraduate computing students increasingly turn to generative AI (GenAI) tools to understand abstract concepts through analogies. Analogies compare an unfamiliar concept to something familiar, but judging whether the comparison holds requires knowledge of both. GenAI may also embed assumptions about who the learner is. GenAI education research centers on output correctness, leaving students' critical reception of analogies largely unexamined. Method: We investigate how students evaluate the accuracy, appropriateness, and assumptions in GenAI-generated analogies, and their perceptions of interest-personalized versus generic technical explanations. Ten students with CS2 experience participated in a pre-survey, a think-aloud task with linked-list and recursion explanations, and a semi-structured interview grounded in the Paul-Elder framework. They judged accuracy, clarity, engagement, and trust separately. Results: Most participants described interest-personalized analogies as more engaging or memorable than generic technical explanations, while trust was mixed. Some trusted the tailored analogies more; others scrutinized them more closely or distrusted the tailoring. Participants with deep source-domain knowledge identified structural flaws requiring that knowledge to recognize. Because personalization and explanation format differed together, these findings do not isolate an effect of personalization alone. Implications: A familiar source flips the student's role. On the concept they are still learners, but on the familiar source they are the expert, and that is the position from which an analogy can be judged. We call this two-sided analogy auditing. GenAI systems should ask what students know, not just what interests them, and treat a flawed analogy as something to inspect and fix rather than accept.
Read the full abstract

Computer Vision

Automatic Red Teaming for Implicit Vulnerabilities of Text-to-Image Models

Red-teaming Text-to-Image (T2I) models is essential for safe deployment, yet it remains particularly challenging against implicit adversarial prompts. Unlike explicit adversarial prompts that can be readily identified and blocked, implicit ones are much harder to detect: the prompts appear benign on the text surface yet still lead to inappropriate visual content. To address this, we propose Adversarial Probing for Implicit VulnErabilities (AdvPIE), a multimodal agentic framework to expose implicit vulnerabilities without requiring access to the parameters of target models. AdvPIE adopts a policy agent to generate and refine implicit adversarial prompts based on the feedback from a judge agent. To construct informative feedback, the judge agent provides modality-specific safety evaluation at both global and relative levels across iterations. To effectively leverage the feedback, we propose a novel Cumulative Adversarial Decoding strategy for the policy agent, which dynamically reweights token distributions to favor tokens that lead to more harmful images while preserving sampling diversity. Extensive experiments on standard and safety-aligned T2I models show that AdvPIE1 effectively uncovers implicit vulnerabilities, outperforming various baseline methods.
Read the full abstract

All authors Chang Ma, Junlin Han, Shuo Chen, Runjia Li, Philip Torr, Jindong Gu

Machine Learning

Connectome-to-Function: Conditional Generative Latent Representations for Reservoir Computing

Connectomes, graph-level maps of neurons and their synaptic connections, provide a structural basis for understanding how brain circuits support function and computation. However, mapping connectome structure to computation remains difficult because these graphs are high-dimensional, sparse, and sensitive to local structural variation. Existing approaches often depend on hand-crafted structural descriptors or task-specific predictors, which limits their ability to represent connectomes in a form that is both generative and functionally meaningful. We propose a conditional generative latent framework that encodes connectome graphs into a compact structural space while using available node-level conditions to guide reconstruction and generation. From this space, the model can reconstruct observed connectivity with a mean edge-reconstruction AUC up to 0.910 and generate new candidate connectomes, enabling a unified analysis of graph structure and computational behavior. Using connectome-derived graphs as recurrent computational substrates, we found that the learned latent space captures functional variation across reservoir-computing experiments, with cross-validated $R^2$ values up to approximately 0.87. Interpretability analysis further revealed task-specific structural mechanisms: in our examples, memory performance is associated with reciprocal recurrent connectivity, whereas prediction and classification are more strongly associated with spectral properties of the recurrent network. These findings suggest an AI-for-science approach to linking neural connectivity to computation and provide a generative and interpretable basis for studying how distinct structural mechanisms shape computational capacity.
Read the full abstract

All authors Zhuolin Yu, Xingyu Liu, Yuanhao Jia, Yunhang Xiao, Hairuo Xue, Feihan Sun, Guozhang Chen

Numerical Analysis

A Second Order Positivity-Preserving Scheme for the Nonlocal Cahn-Hilliard System with Variable Mobility and Logarithmic Flory-Huggins Potential

A second order finite difference scheme is analyzed for the nonlocal Cahn-Hilliard system with variable mobility and the singular Flory-Huggins logarithmic potential. The temporal discretization employs a modified Crank-Nicolson formulation, while the mobility is treated explicitly to guarantee ellipticity and reduce computational cost. To strictly enforce the positivity-preserving property at the discrete level, a nonlinear regularization term is added into the numerical scheme. Rigorous analysis shows that the scheme is uniquely solvable and stable under a modified numerical energy. The well-known analytical challenge associated with the variable mobility has to be handled carefully. To overcome this, a convergence framework is constructed based on two non-standard techniques: (1) a higher-order asymptotic expansion (extending up to the third order in time and fourth order in space) to retain a sufficiently order of accuracy; (2) a two-stage error strategy, wherein a rough error estimate first guarantees the discrete boundedness of the mobility and nonlinear terms, followed by a refined error analysis that yields the optimal convergence rate. Numerical experiments validate the theoretical results and demonstrate the robustness of the proposed scheme.
Read the full abstract

Digital Libraries

VietProfs: A Public Directory of the Vietnamese Academic and Research Diaspora

The Vietnamese academic diaspora spans hundreds of universities and research institutes worldwide, yet there has never been a shared, searchable directory for this community. Prospective graduate students looking for advisors who share their background, researchers seeking collaborators, and organizers looking for speakers have had to rely on manual, one-off web searches. To address this need, we built VietProfs (https://vietprofs.roars.dev), an open, searchable directory of Vietnamese and Vietnamese-diaspora faculty and permanent research scientists outside Vietnam. Each profile pairs a portrait and authoritative diacritic name with verified appointment, degree, and honors data, browsable through a fast client-side web interface. This paper covers two topics: the design of the directory itself and the engineering required to keep it current. Part I describes the directory: its goals, eligibility and appointment track rules, alignment with standard U.S. federal classifications (NCES, NSF, and NIH), profile card design, and current roster insights. Part II reports on automated directory maintenance. Keeping over a thousand records accurate against a constantly shifting web is notoriously difficult. We describe an LLM-assisted maintenance system combining automated search and periodic revalidation with deterministic validation gates, operating under the principle: AI proposes, evidence decides. As of September 2026, the directory includes 1,152 verified records across 492 institutions in 23 countries, covering doctoral cohorts from 1939 to 2026. Aggregate statistics are computed dynamically at runtime, and all changes are audited through public Git history. VietProfs provides a practical community resource while offering a concrete case study in building reliable, agent-maintained public datasets.
Read the full abstract

Astrophysics of Galaxies

FAST Observations of Filamentary and Compact H I Structure in a Magellanic Stream IV Field

The Magellanic Stream (MS) is believed to have formed from gas removed from the Large and Small Magellanic Clouds through tidal forces and hydrodynamic interactions with the Milky Way's gaseous halo. It provides an important laboratory for studying how stripped gas fragments, mixes, and evolves in a circumgalactic environment. In this work, we present HI observations of a $3.8^\circ\times2.2^\circ$ field in the MS IV region, using the data from the Commensal Radio Astronomy FasT Survey (CRAFTS). The total HI mass in the analyzed field is $\simeq 5.3 \times 10^{6} (d/120\,{\rm kpc})^{2}\,M_{\odot}$, where $d$ is the distance to the MS IV gas. The data resolve the emission into three coherent filamentary HI structures with related but distinct velocity trends. To characterize the HI structures and study potential multiphase gas, we adopt a Gaussian decomposition procedure to identify and reconstruct sources. Our results indicate that the field is dominated by one major filamentary HI complex, together with several smaller kinematic clump-like components. Among these identified sources, only one source likely shows multiple velocity components. Because several sources appear spatially overlapped in projection, we further examine their apparent overlap regions using position-velocity (P-V) diagrams. P-V diagrams across the apparent overlap region show no clear intermediate-velocity bridge or V-shaped structure, favoring line-of-sight projection over direct cloud-cloud collision. These results demonstrate the value of deep, high-angular-resolution HI observations for resolving faint emission, compact morphologies, and kinematic structure in the MS.
Read the full abstract

All authors Yuxin Xiao, Qingzheng Yu, Taotao Fang, Di Li, Shulan Yan, Shuo Zhang

Networking

SESO-ISAC: Service-Aware End-to-End Sensing Orchestration for 6G ISAC

Unlike conventional communication services, 6G Integrated Sensing and Communication (ISAC) services require complete end-to-end sensing operations, from data collection to processing, result generation, and exposure. In this article, we propose SESO-ISAC, a service-aware end-to-end sensing orchestration framework that maps a sensing service request to a complete sensing configuration under the sensing execution context. Rather than treating sensing configuration decisions independently, SESO-ISAC captures their interdependence across the end-to-end sensing operation, evaluates candidate configurations according to the applicable QoS requirements, and re- evaluates the selected configuration as the sensing execution context changes. Through two representative case studies, we show that the preferred sensing configuration depends on the overall performance across the applicable QoS dimensions and can change during operation.
Read the full abstract

All authors Youbin Jeon, Laeyoung Kim, Myungjune Youn, Sangheon Pack