Papers

Restoring images degraded by adverse weather remains challenging due to spatially heterogeneous degradations. Many existing weather-specific restoration models rely on weather-agnostic global aggregation, naive cross-scale fusion, and deterministic objectives, which struggle to handle heterogeneous degradations in all-in-one adverse-weather settings. To address these limitations, we propose an Uncertainty-guided Adverse-weather Restoration Network (UAR-Net), a weather-specific AiO framework that integrates a gated transformer with balanced multi-scale skip connections. Specifically, we employ Gated Dual-scale Transformer Blocks (GDTB) to jointly model selective global interactions and multi-scale local structures, a progressive Balanced Multi-scale Skip Connection (BMSC) for balanced multi-scale feature integration, and an Uncertainty-Aware Refinement Head (URH) that performs artifact removal, detail enhancement, and predictive uncertainty estimation. The model is supervised by a Brightness-Aware Energy Loss (BAE-Loss) to encourage accurate reconstruction with well-calibrated uncertainty. Extensive experiments demonstrate that our method achieves state-of-the-art performance across multiple adverse-weather benchmarks. The codes will open source upon acceptance.

In this paper, we consider the Smagorinsky model for the Navier-Stokes equations within a virtual element framework. Under the standard assumption of small data, we prove the existence and uniqueness of a solution. Assuming more regularity to the solutions, we derive the known \del{suboptimal} convergence rates $h$, using $k>1$, for the a priori error estimates of the Smagorinsky model in two dimensional domains. We additionally prove that divergence-free virtual discretizations provide improved convergence orders, with weaker regularity assumptions than in the finite element literature. We conclude the paper with numerical results that corroborate the theory.

Monitoring a DC line with many distributed power converters is a genuinely spatio-temporal problem: the information about a localized fault travels along the whole conductor and reaches a few measurement points mixed with the dynamics of the line itself. This paper develops a two-dimensional geometric Laplace transform (t,x) -> (s_t,s_x) over a commutative subalgebra of the geometric algebra Cl(4,0), isomorphic to Segre's bicomplex numbers, in which two bivectors B_t and B_x act as independent imaginary units for the temporal and the spatial phase. Because the two phases live in algebraically distinguishable planes, a fault at position x_f leaves a transformed residual that factorizes as F_f(s_t) e^{-s_x x_f}: its temporal nature stays in the first factor and its location can be read as a geometric argument of the second. On this representation we build a transmission-line model of the converter line and its space-time dispersion relation, a distributed control by admittance shaping, including an exact treatment of discrete converter sites (spatial sampling, aliasing, and a per-converter droop realization that is exact on the sub-Nyquist band), and a fault diagnosis chain that detects, localizes and classifies injection-loss, shunt, sensor and local-controller faults, extends to multiple simultaneous faults with automatic order selection, and distinguishes the outage of a plant from a cable defect. As an integral object the transform is known in bicomplex analysis, and with a single independent variable it reduces to the complex Laplace transform; the contribution lies in its geometric embedding and in its operational use for fault diagnosis in distributed-converter networks. All results are reproduced by an accompanying open implementation.

Determining the dimension of the $C^1$ cubic spline space $S_3^1(\mathcal{T})$ on an arbitrary nondegenerate planar triangulation has remained unresolved since the 1970s. Schumaker's lower bound includes a local correction $\sigma$ for singular interior four-stars, and it was conjectured that this bound is always attained. We disprove this conjecture by constructing a one-parameter family of nondegenerate realizations of a fixed 18-triangle complex, with only the central vertex moving as $v_6(t)=(t,0)$ on the admissible interval $I=(-3/4,24/55)$. The family exhibits three distinct cases. For $t\in I\setminus\{1/5,3/83\}$, the lower bound is attained and $\dim S_3^1(\mathcal{T}(t))=33$. At $t=3/83$, the central four-star is singular, $\sigma=1$, and the resulting dimension 34 is exactly accounted for by the classical local correction. At $t=1/5$, however, all interior vertices are nonsingular and $\sigma=0$, yet $\dim S_3^1(\mathcal{T}(1/5))=34>P_{\mathcal{T}(1/5)}(1,3)=33$. The smoothing-cofactor calculation shows that the dependence at $t=3/83$ is confined to the central vertex block, whereas the dependence at $t=1/5$ couples all seven interior vertex cycles even though every individual block has full row rank. A complementary Bernstein–Bézier calculation gives the same dimension profile. Thus the singular-four-star correction does not capture every geometry-dependent contribution to $\dim S_3^1(\mathcal{T})$; genuinely global compatibility must also be taken into account.

Broad Learning System is an efficient randomized learning model that expands network width through feature and enhancement nodes and estimates the output weights without deep backpropagation. Its standard least-squares training, however, is vulnerable in two different ways: (i) large residuals caused by noise, outliers, or corrupted labels can dominate the objective, and (ii) all samples are treated as equally reliable even when some lie in ambiguous or locally conflicting regions. This paper proposes IFW-BLS, an Intuitionistic Fuzzy Wave Broad Learning System that addresses these two sources of fragility within one optimization model. The first robustness mechanism is residual-level protection, obtained by replacing the squared loss with the bounded, smooth, and asymmetric wave loss. Boundedness prevents extreme residuals from receiving unbounded influence, while asymmetry allows positive and negative deviations to be penalized differently when the dominant error direction varies. The second mechanism is sample-level credibility control, obtained through intuitionistic fuzzy scores that combine global class-center consistency with local neighborhood conflict. The resulting model evaluates the wave loss on credibility-weighted residuals, so unreliable samples are down-weighted before the bounded loss further limits the effect of extreme errors. A Nesterov accelerated gradient based optimizer is used to solve the proposed objective, avoiding the explicit matrix inversion used in conventional BLS. Experiments on UCI benchmark datasets validate the superiority of the proposed IFW-BLS model over the baseline models; additional corruption experiments also show more stable performance than BLS under noise and outlier contamination.

Large Language Models (LLMs) have recently demonstrated strong capabilities in automated travel itinerary generation. However, real- world travel planning is inherently uncertain: transportation delays, crowd fluctuations, and unexpected stochastic delays frequently inval- idate otherwise feasible schedules. Existing benchmarks like TravelPlanner and TripCraft assume deterministic environments, evaluating only static constraint satisfaction and ignoring whether generated plans remain robust when such uncertainties arise. To address this limitation, we introduce UTP-Bench1 , a large-scale benchmark for uncertainty-aware travel planning. The dataset integrates real-world travel data spanning 504 cities of India, including attractions, restau- rants, accommodations, and multi-modal trans- portation networks. To model realistic disrup- tions, UTP-Bench incorporates empirical delay distributions and crowd-density patterns col- lected from major cities, enabling evaluation of travel plans under stochastic conditions. We further propose three evaluation metrics, namely Buffer Adequacy Score (BAS), Crowd- Aware Timing Score (CATS), and Transport Delay Absorption Score (TDAS), which quan- tify the ability of generated itineraries to main- tain robustness against transit delays and crowd variability. Experiments with state-of-the-art LLMs like GPT-5, Qwen3, Mistral and Phi-4 re- veal substantial gaps between model-generated and human-authored plans, particularly in tem- poral buffering, delay-aware transportation scheduling, and crowd-sensitive planning.

We consider reaction-diffusion eigenproblems with oscillatory diffusion and reaction coefficients. The reaction coefficient magnitude is large: the corrector equation identified by periodic homogenization involves both the diffusion and the reaction operators. We study the numerical approximation of this problem using the Multiscale Finite Element Method (MsFEM). This now classical method is a finite element type method that performs a Galerkin approximation of the oscillatory problem on a specific, problem dependent, basis set. The basis functions are precomputed in an offline stage. Inspired by homogenization theory and using some filtering ideas, we show how to define these basis functions in order to obtain an efficient method. The comprehensive set of numerical experiments that we present, in periodic and non-periodic cases, for the scalar-valued version of the problem (which is then self-adjoint) and for the vector-valued version of the problem (which is then in general non self-adjoint), demonstrates the performance of the approach. Some theoretical arguments complement the numerical observations.

Multi-turn agentic RL increasingly treats credit assignment as a targeting problem: given a terminal verifiable reward, per-turn methods localize credit onto the turns that mattered. We identify the structural quantity that predicts when this is the right move, the verifier information density V_d = k/C (the fraction of an agent's C-step causal chain whose per-turn correctness the verifier exposes), and show that terminal-state verifiers sit deep in a low-V_d regime where targeting is the wrong axis. In controlled shared-rollout comparisons on tau^2-bench that separate reward density from credit geometry, a continuous dense reward spread uniformly beats the sparse binary outcome reward (net-harmful on 4/5 seeds), while concentrating the same advantage on progress turns or on random turns is equally harmful: targeting is second-order. The mechanism is coverage: terminal-state verification collapses the observable signal to a single final-write turn (k=1 in 98% of rollouts) while success requires a 5-8 step chain of prerequisite tool calls. A synthetic phase boundary places the crossover at V_d* ~ 0.8, whereas measured V_d is ~0.15 on tau^2-bench and ~0.4 on BFCL V3; uniform also wins on BFCL, where a matched-concentration shuffled control is negative on 8/8 seeds. The effect reproduces across model families on ToolACE-2-8B (Delta = -0.048 over 32 pre-registered seeds; an independent 20-seed replication is itself significant), and a pre-registered matched-budget breadth sweep traces a monotone dose-response whose deficit vanishes only at full chain coverage, with a reward-to-go arm reaching full-coverage parity. Uniform redistribution is the zero-information coverage default that per-turn schemes must beat; we contribute the matched-concentration shuffled control that any targeting claim should clear.

Multi-turn jailbreak attacks demonstrate that harmful intent can be distributed across dialogue, yet existing methods obscure what conversational mechanisms drive vulnerability. We introduce BLUEPRINT, a safety-evaluation framework separating a factorized social-influence strategy space from WORLDVIEWSIM, a cross-turn situational context module. Monte Carlo Tree Search optimizes turn-level combinations of 18 theory-grounded influence factors across a four-turn trajectory. Across six frontier models, BLUEPRINT achieves near-ceiling ASR on major open-weight and proprietary models, while requiring the fewest average queries (2.46). The resulting trajectories further reveal model-specific vulnerability among resistant targets: each responds to distinct influence factors and strategy transitions, yet all share a common recovery pathway-shifting toward concrete, executable task framing consistently escapes hard-refusal states. Ablations confirm operational cues matter most: making requests actionable has the largest impact, gain framing is unusually potent, and some legitimacy appeals can backfire. These findings suggest robust multi-turn safety requires monitoring not only harmful content, but also how dialogue state makes unsafe requests appear concrete and locally executable.

Multi-axis 3D printing enables support-free fabrication and improved part quality, but robustly processing real-world geometries remains challenging. Models from design workflows or direct data acquisition often contain solid–shell combinations and non-manifold structures. Handling such models in the wild typically requires time-consuming geometry repair, which may alter the intended geometry. In this work, we present WildFab, a computational framework for multi-axis 3D printing that directly computes spatial toolpath and global collision-free motion from input models. Our pipeline builds on a hybrid query representation that combines a neural unsigned distance field (UDF) with a regularized generalized winding number field (reg-GWN). The UDF supplies differentiable surface-distance and direction queries, while the reg-GWN resolves near-surface ambiguity in the fitted UDF by providing reliable surface localization and a solid-void indicator. Based on this representation, we introduce a high-precision spatial toolpath computation algorithm that iteratively projects points between optimized guidance-field level sets and reg-GWN gradient-magnitude ridges. Subsequently, we develop an efficient and robust coarse-to-fine collision checking scheme for motion planning: UDF-based rejection first identifies potential collisions, while time-varying reg-GWN verification accurately resolves collision pairs for both solid and shell components. We validate WildFab on diverse inputs, demonstrating successful computation from non-manifold parametric surfaces, voxelized topology-optimization results, implicit models, raw scanned point clouds, and non-watertight meshes. The fabrication results highlight our method's ability to advance end-to-end design-to-3DP workflows.

A capable brain-MRI report generator can still be, in effect, diagnostically silent. When a multi-chain chain-of-thought (CoT) reporter built on a medical Mistral-7B backbone is evaluated on held-out cohorts, it names most meningiomas and almost all metastases "glioma" (diagnosis recall 0.44/0.07). Yet the answer is not absent from the model: a supervised linear probe applied to its frozen segmentation features recovers the three tumour cohorts at 0.82 macro-F$_1$ (5-fold cross-validation; chance $\approx$0.33). We introduce NeuroFusion, an assistive reporter that surfaces this latent signal rather than overriding it: discriminative field-classifier heads over per-lesion features condition a fast, single-pass draft-then-review decoder on their committed outputs. Built on the identical Mistral backbone, this restores the diagnosis (meningioma 0.92, metastasis 0.75) and wins 8 of 9 prose-content comparisons across three held-out cohorts (RaTEScore, RadGraph-F$_1$, GREEN; Holm-corrected paired BCa), with no significant loss on the ninth, at 5-6x lower latency ($\approx$80 vs. 457 s/case). A controlled negative result sharpens the mechanism: a learned diagnosis pin that overrides the decoder instead of merely informing it collapses out-of-distribution metastasis recall to 0.03. Grammar-constrained decoding keeps 92.3% of records schema-valid, making every sentence entailment-checkable (7.5% contradicted vs. 36.8% for the direct baseline). In a blinded nine-case pilot, two board-certified neurologists independently rated NeuroFusion highest in every tumour type, the only system with zero critical errors, and gave it the top-rated sign-off in eight of nine cases (six outright, two ties).

Air conditioner (AC) loads offer promising flexibility for active distribution networks to manage uncertainties, such as those in renewable energy generation, electricity prices, and load demand. However, real-time scheduling of ACs is challenging due to their massive temporal coupling constraints and time-causal uncertainties. To address this, a novel time-causal aggregation-based approximate dynamic programming (TCA-ADP) algorithm is proposed for efficient scheduling. The time-causality requirements for aggregating state variables are first analyzed to align with the real-time sequential decision-making process. Subsequently, an enhanced aggregation model is developed to ensure both high accuracy and adherence to time causality. The aggregation process is further reformulated as a linear program to optimize aggregation parameters and enable tractable computation. Accordingly, the TCA-ADP leverages aggregated state variables to approximate the value function as a new way, balancing computational efficiency and economy against the large value function space of massive ACs. By training the value function offline using historical data, the TCA-ADP efficiently achieves near-optimal real-time scheduling of massive ACs through parallel and closed-form disaggregation. Case studies demonstrate the effectiveness and scalability of the TCA-ADP, highlighting its aggregation accuracy, uncertainty handling, and the trade-off between economy and tractability.

Heating, ventilation, and air conditioning (HVAC) loads, with their rapid response capabilities, can provide considerable intra-hour flexibility on the demand side for reserve provision in order to follow the fast variations of renewables. However, scheduling massive HVACs is challenging due to computation complexity and the uncertainty of outdoor temperature. In this paper, we first introduce a novel continuous-time (CT) aggregation model to reveal the potential intra-hour flexibility of HVACs. For accurate aggregation, a new affine transformation is designed to handle the heterogeneity in high-dimensional feasible region. Further, for reliable aggregation in practical environment, the outdoor temperature uncertainty is constructed by distributionally robust chance constrains and integrated into the aggregation model. Secondly, for the tractable calculation of the proposed CT aggregation model, a cascade of tailored reformulation techniques is proposed, including the Bernstein polynomial spline, polytope projection, and linearization transformation. Thirdly, a customized hierarchical dispatch framework is proposed via incorporating the proposed CT aggregation model into reserve provision in power system dispatch, so as to efficiently schedule massive HVACs to cope with the renewable uncertainty. Case studies verify the effectiveness and scalability of the proposed CT aggregation model in aggregation accuracy, intra-hour flexibility utilization, and uncertainty handling.

Sparse mixture-of-experts (MoE) models use an independently parameterized router at each sparse layer to select experts for every token. Prior work has shown that routing decisions across depth can often be predicted from earlier routing signals, suggesting that routing is not fully independent across layers. However, the structure behind this predictability remains unclear. In this work, we provide evidence that routing-relevant states across layers share a common geometric structure that is obscured by layer-specific coordinate systems. We isolate the control subspace of each router and align these spaces into a shared canonical representation using generalized orthogonal Procrustes analysis. After alignment, a single linear transition reaches $R^2=0.39$–$0.71$ and retains 79–90% of the predictive power of separately fitted layer-specific dynamics, indicating that much of routing-state evolution follows a reusable process across depth. We then ask whether this shared dynamics is specific to routing or simply reflects the smooth evolution of hidden representations. A matched-rank comparison shows that residual representations are often easier to predict across layers, while router-control states preserve the model's expert choices much more faithfully. This separates generic cross-layer predictability from routing-specific information. Finally, we test whether the predicted canonical states remain meaningful when used in place of native routing states. The transported states preserve local routing behavior, while learned state evolution reduces $\Delta\mathrm{NLL}$ relative to simple persistence by 15.7% on OLMoE and 6.2% over a 10-router horizon on Phi.

Conventional task-level evaluation asks whether a robot policy completes a specified action, but can miss failures that emerge only during physical human contact. This limitation is critical in contact-rich assistive tasks, where meaningful evaluation requires a physically responsive human, interaction-quality assessment beyond task success, and a leak-free observer-scorer protocol. We introduce a physics-consistent benchmark for contact-rich human-robot interaction, instantiated in robot-assisted bathing. The benchmark combines a deformable, passively responding human, physics-aware scores alongside task-level success, and a frozen vision-only / scorer-only evaluation protocol. To establish physical validity, region-wise simulated responses are calibrated against force-indentation measurements from Franka impedance pushes on a medical-care manikin. Under a frozen T1-T7 protocol with 140 runs per method, an LLM-augmented state machine (State Machine) achieves 72.9% task success but drops to 56.4% after correct-region and force-safety screening; VoxPoser produces lighter and more stable contact but completes only 27.9% of trials; and zero-shot pi0.5 achieves 0.7% task success with no correct-region or safety-gated successes. These results show that task completion alone does not imply physically valid contact and motivate physics-aware screening before deployment of contact-rich assistive robot policies.

Autoregressive vision language models unify heterogeneous perception tasks but are highly susceptible to compounding errors. On-policy distillation (OPD) bridges the training-inference mismatch by training students on their own rollouts. However, unreliable student predictions, especially early in training, can derail the trajectory and degrade the quality of teacher supervision. While recent interleaved distillation methods allow the teacher to verify and replace student tokens, they primarily rely on rigid ranking metrics rather than exact teacher confidence, and they overlook how intervention decisions can inform token-level supervision. To address this, we introduce Confidence-Aware On-Policy Distillation (CA-OPD), a framework that couples reliable rollout construction with adaptive supervision. CA-OPD utilizes teacher confidence to selectively correct unreliable student transitions, gradually transferring rollout control to the student via a strict-to-relaxed schedule. Crucially, CA-OPD aligns knowledge transfer with these intervention decisions: corrected positions receive direct cross-entropy supervision from the teacher's prediction, while retained positions benefit from the teacher's full predictive distribution. Evaluated in a multi-teacher setting for GUI grounding and optical character recognition, CA-OPD substantially improves the Qwen3.5-0.8B baseline across all six target benchmarks, including gains of $9.50$ points on ScreenSpot-Pro and $6.72$ points on OCRBench-v2 English. Controlled studies further show that the gains depend on intervention placement, progressive rollout control, and intervention-aligned supervision, rather than intervention frequency alone.

Argumentation frameworks are useful tools for representing and reasoning with information in a variety of settings, e.g. in supplementing AI models as they perform classification tasks, with a notable benefit of providing additional explainability. In this paper, we introduce contrastive explanations for Quantitative Bipolar Argumentation Frameworks (QBAFs), one such formalism. Unlike most existing explanations for QBAFs, which explain the reasoning outcome of a single argument of interest (i.e. a topic argument), contrastive explanations explain the difference between two topic arguments. We introduce a general form of contrastive attribution functions (CAFs) and establish a set of general properties they should satisfy. We introduce CAFs based on removal, gradients and Shapley-values, and study their properties. Finally, to illustrate contrastive explanations, we demonstrate their usefulness in healthcare and bias identification settings.

Radiology reports are written primarily for clinicians, and their specialized terminology often makes them difficult for patients to interpret. As a result, many patients turn to publicly available Large Language Models (LLMs) to help explain their reports, despite well-documented risks of factual inaccuracies and hallucinations. Automated lay-summary generation has emerged as a promising alternative, yet the effectiveness of retrieval-enhanced and clinically informed approaches for radiology-specific communication remains underexplored. This study investigates the extent to which Retrieval-Augmented Generation (RAG) and Named Entity Recognition (NER) improve the quality, factual consistency, and readability of automatically generated lay summaries compared with standard LLM-based generation. We develop a framework combining NER-based extraction of clinically relevant findings with a RAG mechanism for contextual grounding, evaluated across few-shot and fine-tuned variants of two models (Qwen, BioBART). Results show that NER consistently improves readability and overall quality, while RAG alone offers no benefit and can introduce hallucinations from irrelevant retrieved terms. Combining RAG with NER degrades performance in few-shot settings but improves readability when fine-tuned. Fine-tuned BioBART with NER achieves the best overall performance, highlighting entity-aware extraction as the primary driver of improved patient-friendly summaries.

Vision-language models (VLMs) can solve visual CAPTCHAs without task-specific training, but the agents built on them approach every challenge from scratch. For such an agent, the hundredth instance of a familiar puzzle costs as much time and compute as the first. Specialized detectors invert the trade-off, answering in milliseconds but only for categories they were trained on. Neither improves with exposure. We study what changes when a solver improves with use. Our system pairs a fine-tuned YOLOv8 detector with an open-weight VLM behind a confidence-based router, and runs entirely from screenshots and operating-system input events, with no browser automation or DOM access. It reaches 85.4% overall and 84.2% macro accuracy across 16 classes, exceeding either component alone. Every answer VLM produces also serves as a training label, so the detector absorbs categories it was never trained for, typically after one or two encounters and without human annotation. The same loop also repairs it. A CAPTCHA operator can perturb images against the publicly released detector and drive its accuracy to 0%, but the perturbations leave VLM untouched, and its labels let the detector recover. Under a year-long simulated arms race in which the CAPTCHA operator re-crafts its perturbations each month, the solver recovers every round, and a cheap ~70%-accurate open-weight teacher hardens it as effectively as a perfect oracle. Visual CAPTCHA defenses that assume a failing bot stays failing therefore understate how quickly an adaptive solver returns.

Political evasion refers to responses that engage with a question while withholding the requested information. Recent NLP work frames political evasion as a classification task using a two-level taxonomy of response clarity and fine-grained evasion strategies. Existing work on response clarity and evasion classification is limited to English, leaving open whether the taxonomy and model behavior transfer across languages and political contexts. We introduce PolERo, a dataset of 3,574 human-annotated question-answer pairs extracted from official transcripts of five Romanian presidents. We evaluate multiple classification approaches on both datasets under matched conditions, including TF-IDF baselines, fine-tuned encoder models, a proposed sliding-window encoder, and zero/few-shot LLM prompting. We study cross-lingual transfer through joint bilingual training and machine-translation-based data augmentation. Our results indicate that fine-tuned encoders are competitive, cross-lingual transfer is asymmetric, and ambivalent evasion categories involving pragmatic cues remain the main challenge across all model families.

Sampling error yields exclusively reactive, non-lesional brain parenchyma in a significant proportion of intracranial biopsies, leaving the underlying disease undiagnosed. We benchmark four pathology foundation models (UNI2-h, Virchow2, Prov-GigaPath, H-optimus-0) as frozen patch encoders within a shared attention-based multiple-instance learning framework using 245 whole-slide images from 186 patients with confirmed downstream diagnoses. We first show that coarse disease-category prediction can be reproduced largely from slide size alone. After restricting classification to three finer diagnostic distinctions within common tissue categories, this confound no longer explains performance, yet disease labels remain predictable above chance under permutation testing (p $\le 10^{-4}$ throughout). Surprisingly, performance is statistically indistinguishable across all foundation-model encoders, suggesting that recovering these weak morphological signatures is not limited by current patch representations. Signed instance-contribution maps and expert review further test whether predictive evidence localizes to reactive parenchyma rather than sampling-induced bias like blood introduced during tissue sampling. These results position acquisition-shortcut auditing via a provenance-only baseline as a necessary control in computational-pathology benchmarks, and show, once that confound is removed, that weakly supervised models still recover disease signal from tissue conventionally regarded as non-diagnostic.

An edge card of a graph is obtained by deleting one edge, and a class edge-reconstruction number asks for the fewest carefully selected cards that identify the graph when its class is known. We determine the sharp universal bound for maximal planar graphs. Two selected cards always suffice, and the octahedral graph shows that two can be necessary; some maximal planar graphs are already identified by one card. The argument exploits the fact that deleting a flippable edge leaves a single quadrilateral whose two diagonals give the only possible maximal-planar completions. Degree information then rules out the competing completion, with a separate argument for graphs containing a vertex of degree three. This settles a problem posed in a 2010 survey on reconstruction numbers.

We study the low-degree Steiner forest decomposition. Given a graph $G=(V,E)$ and a terminal set $U\subseteq V$, the standard decomposition returns a set $X\subseteq V$ of size at most $|U|/2$ and a forest $T\subseteq G-X$ of maximum degree $\Delta$ such that, for every connected component $C$ of $G-X$, some connected component of $T$ contains all terminals in $U\cap V(C)$. This is the central decomposition behind several connectivity oracles under vertex failures [DP20, LS22, LW24]. The state-of-the-art algorithms either take $O(mn\log n)$ time with degree bound $4$ [DP20], or take $m^{1+o(1)}$ time with the weaker degree bound $O(\log^{2}n)$ [LW24]. We show that if $T$ is allowed to contain vertices of $X$, then a degree-$4$ decomposition can be computed by a very simple algorithm in $O(m\alpha(n))$ time. Further, we show that this relaxed decomposition is equally useful for constructing connectivity oracles under vertex failures. As a consequence, we obtain a deterministic connectivity oracle under $d$ vertex failures with $\tilde{O}(m)$ space, $\tilde{O}(md_\star)$ preprocessing time ($d_\star$ is an upper bound on the number of failed vertices), $\tilde{O}(d^{2})$ update time, and $O(d)$ query time. Up to polylogarithmic factors, this oracle strictly improves all known oracles; in particular, it removes the $n^{o(1)}$ factors from the preprocessing and update times of [LS22, LW24].

Multiobjective evolutionary algorithms (MOEAs) naturally expose population-level parallelism, but many mature implementations encode their computation in sequential program structures designed for central processing units. Exploiting modern tensor computing platforms therefore requires more than direct code translation: the implementation must be restructured without changing the defining optimization mechanism of the underlying MOEA. We formulate automatic tensorization for MOEAs as semantics-guided computational restructuring and develop Evolutionary Code Conversion (EvoCoCo), a multi-agent framework that realizes this formulation. EvoCoCo reconstructs algorithm-specific states, dependencies, operators, and update logic into a structured semantic representation and organizes them through a shared tensorization blueprint. Specialized transformation branches then explore alternative tensor realizations, while execution feedback guides validation, repair, and candidate selection. Experiments on a benchmark of 48 MOEAs evaluate migration reliability, optimization fidelity, and computational scalability. Under matched large language model backends, EvoCoCo attains higher migration reliability than direct one-shot translation. Across the benchmark suites, 88.2% of valid comparisons satisfy the predefined optimization-fidelity criterion. The tensorized implementations also exhibit increasing acceleration on graphics processing units as population size or decision dimension grows, with median measured speedups ranging from $22.6\times$ under population scaling to $80.2\times$ under decision-dimension scaling. External-source and ablation studies further assess transfer beyond the main benchmark and the roles of the major framework components.

Gaussian graphical models (GGMs) are essential tools for interpretable structure learning. However, in high-dimensional, small-sample regimes, the available data is often insufficient for the maximum likelihood estimator to exist. Colored Gaussian graphical models (CGGMs) mitigate this limitation by imposing symmetry constraints through graph coloring, which reduces the required sample size. This minimal number of observations needed to guarantee that the estimator exists almost surely is defined as the maximum likelihood threshold (MLT). Here, we address the computation of the MLT for CGGMs by focusing on its geometric formulation: finding the minimum rank of a sample covariance matrix such that its projection lies almost surely within the interior of the cone of sufficient statistics. We establish a unified theoretical framework, extending results from uncolored to colored models and introducing new symbolic algorithms. Furthermore, we present a computational study integrating sampling with topological data analysis (TDA) to investigate the local geometry of the cone of sufficient statistics. Our results demonstrate the potential of TDA to overcome the computational bottlenecks of traditional symbolic algebraic methods, particularly Groebner basis computations, in analyzing the likelihood geometry of CGGMs.

While existing work on LLM authorship attribution (AA) has made progress, available benchmarks remain limited, often focusing on English, controlled settings, or relatively outdated models, with the few multilingual studies considering only relatively short texts. We introduce MultiGhostBench, a multilingual benchmark comprising 928 books generated by five recent LLMs across six languages and three scripts, with an average length of approximately 59K words per book. The benchmark supports evaluation under domain, author, and language shifts. Evaluation of representative AA methods shows that no single method consistently performs best across settings, and performance generally degrades under distribution shifts. Transformer-based detectors can retain generator-related information across languages, although transfer effectiveness varies by language pair, whereas statistical and fingerprint-based detectors are more language-dependent. We envision MultiGhostBench as a valuable resource for the development and evaluation of robust AA methods. The dataset and code can be found at https://github.com/GrecoMT/MultiGhostBench.

High-resolution Synthetic Aperture Radar (SAR) imagery is critical for precision analysis such as automatic target recognition, yet its acquisition is costly. Although generative image super-resolution (ISR) models offer a promising alternative, current smooth-approximation based diffusion frameworks often struggle to preserve the coherent scattering statistics, causing stochastic structural distortions that are less consistent with real SAR physics. To address this, we propose Semantic Prototype-Guided Super-Resolution (ProSR), reformulating SAR ISR as a semantically-guided discrete token prediction task within a quantized latent space. By mapping signal features to discrete scattering primitives, ProSR preserves the impulsive nature of SAR without over-smoothing. Furthermore, we integrate a Self-Supervised Learning backbone into SAR ISR to extract label-free semantic priors, overcoming label scarcity. Guided by these priors, we introduce Semantic-Aligned Detail Encoding to decouple high-frequency signals into discrete scattering primitives. In parallel, the Semantic Prototype Map Generator explicitly constructs semantic prototype maps, allowing Prototype-Map-Guided Attention to route the information flows within identical categories and mitigate inter-class interference. To validate our approach, we present a large-scale 0.25m resolution benchmark from the Umbra Open Dataset. Experimental results show ProSR achieves superior visual quality while preserving essential scattering characteristics required for practical SAR applications.

Acoustic sensing can monitor the daily activities of older adults without cameras or wearables, yet the possibility of capturing private conversations remains a barrier to its deployment. The engineering application addressed in this paper is privacy-preserving acoustic monitoring of daily activities in assisted living. Our contribution in artificial intelligence is a speech-removal front end: a U-Net that masks the short-time Fourier transform of ambient audio and returns a waveform in which speech is suppressed and the sounds needed for activity recognition are kept. The model is trained only on synthetic mixtures of public sounds and speech, so no speech has to be recorded in the target homes. We evaluate it with group-disjoint cross-validation on environmental and domestic sound datasets mixed with reverberant speech, on speech from unseen corpora and on real home recordings, and measure privacy with three voice activity detectors, Whisper speech recognition and a speaker-verification attack. With speech 5 dB above the background, the macro-averaged F1 score of activity recognition is 0.574 and 0.661 after removal on the two datasets, compared with 0.233 and 0.178 without processing, and higher than that of speech enhancement, speech separation and detector-gated muting at every speech level. Whisper recovers 5.2% of the spoken words, compared with 89.0% without processing and a chance floor of 2.2%. On the real recordings, the speech proportion estimated by a voice activity detector falls from 8.5% to 2.7%. The method reduces, but does not eliminate, spoken content and speaker identity.

We study the dynamics of Stochastic Gradient Descent (SGD), which is known to steer deep neural networks toward invariant sets that correspond to simpler subnetworks. How this steering unfolds over time remains poorly understood. We answer this by modeling the stochastic gradient flow (SGF) as a percolation process, in which nested architectural symmetries force subnetworks to merge in discrete blocks rather than by single-edge attachment. These structural transitions register as variance spikes in a macroscopic order parameter echoing physical phase transitions. We further state sufficient conditions under which the trapping argument carries over to Adam and AdamW under heavy-tailed gradient noise and measure them on a trained Transformer.

With the proliferation of LLM agents, the ability to understand and diagnose failures in agents is essential to achieving superior effectiveness and trustworthiness. As agent failures often manifest via long and complex trajectories, manually finding the needles in the haystack is untenable. However, traditional diagnosis techniques for software bugs can hardly address LLM agent failures, while completely relying on LLMs as the judge yields unreliable diagnosis results. To overcome these challenges, this paper presents AGENTSCOPE, a new neuro-symbolic approach for agent failure mode diagnosis. The key principle of AGENTSCOPE is to abstract agent behavior, based on its trajectories, into structured representations. Furthermore, AGENTSCOPE introduces the concept of neural invariants to specify agent behavior properties. AGENTSCOPE leverages LLM-guided reasoning atop the structured representation against neural invariants to pinpoint both the failure step and its type in the trajectory. We show the effectiveness of AGENTSCOPE on publicly available agent failure datasets (Who&When) and a more comprehensive dataset created by us (AgentErrata), where AGENTSCOPE significantly outperforms the current state of the art in fault localization and attribution accuracy. Our work shows that integrating structured abstractions with LLM-guided reasoning enables effective, reliable, and interpretable diagnosis for agent failures.