Computer Vision
Papers
Fresh from arXiv, typeset properly. Highlighted words trade on the market.
All of arXiv
Thursday 10 September
We present a deep-learning pipeline for enhancing the detection of faint moving objects in optical space situational awareness (SSA) imagery through automated star removal and background reconstruction. Detecting low signal-to-noise ratio (SNR) objects remains extremely challenging in optical observations, particularly in the cislunar (X-GEO) environment, where structured sky backgrounds, dense stellar fields, and scattered moonlight significantly degrade the performance of classical detection algorithms. To address this problem, the proposed pipeline combines a lightweight segmentation network (Tiny-U-Net) to generate stellar masks with a partial-convolution variational autoencoder (astro-VAE), designed to learn the statistical distribution of astronomical backgrounds and perform context-aware inpainting of masked regions. The reconstructed background maps can then be used as a preprocessing step to suppress fixed sources and background inhomogeneities prior to detection. As a proof of concept, the approach is integrated with a shift-and-stack scheme and evaluated on real ground-based telescope observations targeting the X-GEO region. Results demonstrate that the method reconstructs star-free backgrounds with high fidelity, while preserving moving targets and significantly enhancing detectability, thereby providing an effective data-driven preprocessing strategy for faint moving-object detection in optical SSA scenarios.
Adaptation and Self-Organizing Systems
Reconstructing the information processing capacity of physical systems from noisy observations
Driven dynamical systems can compute when their transient states encode complex transformations of past inputs. The information processing capacity (IPC) framework allows for a detailed accounting of these computational properties, however its interpretation in noisy systems has remained incomplete. In this work, we clarify how noise affects the IPC and how one can reconstruct the noiseless IPC. First, we show how to distinguish the dynamics of an unperturbed system from the noise-free component of the stochastic dynamics: The IPC measured for responses averaged over noise realizations is in general not the same as the IPC of the unperturbed system. We explicitly demonstrate that noise can redistribute computational capacity and sometimes even enhance performance on particular tasks, rather than merely degrading a fixed computation. We then introduce covariance reconstruction by orthogonal projection (CROP), which reconstructs the covariance and IPC of the noise-free component directly from noisy observations, without requiring a detailed model of either the system or the noise. At fixed total measurement budget, numerical tests on a classical nonlinear reservoir show that CROP estimates the noise-free IPC more accurately than the standard practice of ensemble averaging over repeated trials. We find the same advantage in a quantum reservoir subject to unavoidable measurement noise. Our results provide a general route to recovering the computational structure of noisy physical systems from finite observations.
Read the full abstract
History and Philosophy of Physics
Scientific Promise
Scientists constantly face decisions about what lines of research to pursue. This Element introduces the philosophical debate about scientific pursuitworthiness. It explains how it can be rational to pursue a theory even if the theory is less well supported than its rivals, and it discusses existing philosophical frameworks for guiding pursuit decisions. The Element also develops a new perspective. Existing accounts focus predominantly on theories, while experiments are largely neglected. This is an important shortcoming. Theoretical promise depends on experimental promise, and experimental promise raises questions of its own. Drawing on the epistemology of experimentation, the Element advances an account of experimental pursuitworthiness. It is argued that such pursuitworthiness depends on experimental virtues like a clear signal and simplicity of design. Moreover, the kinds of uncertainty that constrain the assessment of scientific pursuits are examined. Finally, the Element highlights open questions in the philosophy of scientific pursuitworthiness.
Read the full abstract
Number Theory
Log-crystalline representations and $(\varphi, \Gamma)$-modules
Let $F$ be an absolutely unramified mixed characteristic local field with rings of integers $O_F$. For a small affine algebra $R$ over $O_F$, we consider the logarithmic étale fundamental group $G_R$ of its generic fibre equipped with a "horizontal" log-structure, in the sense of Fujiwara–Kato. We study $p$-adic representations of $G_R$ and obtain a classification of these representations in terms of étale $(\varphi, \Gamma_R)$-modules. Moreover, we define and study the notion of log-crystalline representations of $G_R$, a generalisation of crystalline representations from the non-logarithmic/smooth case. Furthermore, in the logarithmic setting, we show that log-crystalline representations of $G_R$ are equivalent to Wach modules for $R$, extending our previous results from the (non-logarithmic) crystalline case.
Read the full abstract
Computer Vision
Uncertainty DMD: Restoring Diversity in Few-Step Autoregressive Video Distillation
Few-step distillation improves the efficiency of autoregressive (AR) video generation, but often causes diversity collapse: under the same prompt, different noise samples tend to produce highly similar videos with weakened motion dynamics. We analyze this degradation in Distribution Matching Distillation (DMD)-distilled AR video generators and find that, in the autoregressive setting, it takes the form of a structured uncertainty collapse: the mode-seeking bias of DMD maps different noise samples to nearly identical first chunks, and the deterministic AR cache then propagates this collapsed state to all subsequent chunks, turning a local loss of stochasticity at the rollout root into a global suppression of temporal variation. Based on this analysis, we propose Uncertainty DMD, a simple uncertainty-injection framework that restores stochasticity at two key stages of AR generation: a timestep perturbation for the first chunk to increase first-chunk diversity, and a stochastic cache-writing mechanism for later chunks to preserve uncertainty in autoregressive conditioning. The method requires no architectural changes and introduces only lightweight perturbation operations. The same perturbation mechanisms are used during both training and inference. Experiments show that Uncertainty DMD consistently improves diversity and motion dynamics while maintaining comparable per-sample visual quality.
Read the full abstract
Distributed Computing
Can AI Remediate Backend Failures Safely? GuardedAct with Blast-Radius-Aware Sandboxing
Large Language Models (LLMs) have shown promising capabilities in generating remediation actions for microservice failures. However, directly executing AI-generated repair actions in production risks cascading collateral damage. We propose GuardedAct, a sandbox-first remediation framework that interposes a blast-radius-aware verification layer between the LLM action generator and the production environment. GuardedAct operates in four phases: (1) ingesting a diagnosis report together with the live system topology and recent telemetry, (2) prompting an LLM to produce a ranked list of candidate remediation actions, (3) simulating each action in a lightweight digital-twin sandbox that estimates the blast radius and assigns a risk label, and (4) enforcing a rollback-confidence gate that auto-executes only low-risk actions while escalating high-risk ones for human review. We evaluate GuardedAct on five fault scenarios injected into the DeathStarBench social-network application. Experimental results show that GuardedAct achieves an overall recovery rate of 87.4% while reducing collateral damage by 79.7% relative to direct LLM execution (from 25.6% to 5.2%), at the cost of a modest sandbox-induced increase in mean time to recovery (approximately 8 s). Ablation studies confirm that each component contributes meaningfully to the safety-speed trade-off.
Read the full abstract
Signal Processing
Exact Bayesian Tracking of Dynamic Network Topologies
Tracking the temporal evolution of network topologies is a fundamental challenge in social networks, epidemiology, and sensor systems, among others. This paper develops an exact Bayesian tracker for unweighted, directed graphs using nodal signal observations. This framework yields the full posterior probability distribution over network states at each time step, naturally enabling uncertainty quantification, prediction, and principled decision-making. We model the network dynamics as a Markov process on the Boolean hypercube, where edges transition independently according to a flip probability. For efficient computation, we cast the prediction step as a dyadic convolution, and leverage the Fast Walsh-Hadamard Transform to reduce the computational cost from $\mathcal{O} (4^k)$ to $\mathcal{O} (k 2^k)$, where $k$ is the maximum node degree. When the network transition probabilities are unknown, we develop an Expectation-Maximization framework to learn them from the observed signals. Comprehensive experiments on synthetic and six real-world datasets validate the proposed method and demonstrate its superior tracking accuracy, faster recovery from topological changes, and meaningful uncertainty estimates compared to state-of-the-art and classical baselines.
Read the full abstract
Artificial Intelligence
AI-Powered Flare Combustion Efficiency Estimation
Achieving high combustion efficiency in flare stacks is crucial for adhering to regulatory standards and controlling the release of hydrocarbons into the environment. Traditional instruments like gas analyzers and hyperspectral cameras are expensive, fragile, and require frequent calibration, which makes them impractical for remote or budget constrained industrial sites. We propose an innovative solution that combines a lightweight vision-language encoder with a compact multi-layer perceptron to predict combustion efficiency directly from low-cost thermal video footage. The fully trained model is integrated into an easy-to-deploy graphical user interface. This interface overlays predicted combustion efficiency values on each video frame, displays real-time trends in combustion efficiency, shows the distribution of combustion efficiency across all frames in the video, and allows users to export CSV reports. Over a six-month period, the system achieved 99% uptime and required less than 15 minutes of maintenance per week.
Read the full abstract
Digital Libraries
INDRA: A New AI Tool for Exploring Tobacco, Fossil Fuel, and Chemical Industry Archives
Five decades of litigation have disgorged hundreds of millions of pages of formerly secret business records from the tobacco industry, along with documents from the makers of drugs, chemicals, food, firearms, and fossil fuels. Yet these archives have been effectively inaccessible to general-purpose large language models (LLMs) because they have never been compiled into an LLM-readable corpus. Chatbots may be familiar with some of the materials contained in such archives but, with no direct access to the documents, they are vulnerable to hallucination and other defects. Here we introduce INDRA, a research platform designed to remedy such failures by embedding the conventions of archival historiography into a system-level protocol governing every output. The platform federates UCSF's Industry Documents Library, Columbia and CUNY's ToxicDocs, Stanford's SRITA, and other heretofore siloed collections, and provides three interlinked safeguards: (1) a closed evidentiary sandbox confines the model to a user-selected corpus, blocking retrieval from external sources that could introduce bias; (2) real-time provenance tagging marks the boundary between archival evidence and parametric inference; and (3) a system-level protocol enforced by deterministic scripts guides the structure of every output. Together these safeguards prevent the model from conflating "the documents say X" with "I think X" or "I learned X from prior training." The result is an LLM-powered research partner enabling massive multi-archival investigations, a tool whose outputs are designed to be checked rather than trusted, and whose architecture makes the conditions of knowledge production visible and auditable. Three case studies demonstrate the method's analytical value and limitations, including what we call the Heraclitus effect, the steppingstone dilemma, and the gullibility (or mafia) problem.
Read the full abstract
Audio and Speech Processing
Preference Optimization with LALM Feedback for Continuous Autoregressive Non-Verbal Vocalization Generation
We propose a preference optimization framework with Large Audio-Language Model (LALM) feedback for controllable non-verbal vocalization (NVV) generation in continuous autoregressive speech models. To construct preference data without human preference annotation, we build a bilingual prompt corpus by combining NVV-injected real transcripts with LLM-generated semantically aligned prompts, perform stochastic model rollouts, and use a LALM to rank candidate utterances and form same-prompt chosen–rejected pairs. We then adopt a two-stage optimization strategy: Rejection Sampling Fine-Tuning (RSFT) first adapts the model to LALM-selected high-scoring samples, followed by Anchored Flow-DPO, which formulates pairwise preference optimization using utterance-level flow-matching loss and retains the chosen-sample flow-matching objective as an SFT anchor. This design enables DPO-style preference learning without explicit sequence likelihoods while preserving direct supervision on preferred realizations. On the official 1,600-utterance NVVSpeech Challenge Track~2 test set, our method achieves a Final Track2Score of 75.80 (79.39 ZH / 72.21 EN), outperforming the VoxCPM2 baseline by +1.84. The improvements are mainly driven by higher NVV Accuracy and NVV Perceptual Effect, while Overall Quality remains stable.
Read the full abstract
Numerical Analysis
A note on the convergence analysis of Laguerre approximations for analytic functions
In a recent paper (H. Wang, Math. Comp. 93, 2861-2884, 2024), Wang has presented a number of results concerning the convergence of approximations based on generalised Laguerre polynomials, claiming to have provided "the first rigorous proof of root-exponential convergence of Laguerre approximations for analytic functions". In this note we argue that the proofs of the main results of Wang's paper are incomplete, because they rely on taking a limit under an integral sign, and this is not properly justified. We explain how the proofs could be completed using the dominated convergence theorem, provided that certain nontrivial inequalities involving generalised Laguerre polynomials and their weighted Cauchy transforms hold. We make a related conjecture, which, if true, would complete the proofs of Wang's results. This conjecture remains unproven, but appears plausible, based on numerical experiments.
Read the full abstract
Computers and Society
SoulAuth: An Actor-native Identity Architecture and Rust Reference Implementation for Humans and Long-lived AI Actors
As AI systems move from transient model invocations toward long-lived actors that persist across credentials, clients, sessions, and runtime instances, identity infrastructure must answer a basic question: where should the canonical continuity boundary be placed? This paper introduces Actor-native Identity and presents SoulAuth, an open-source Rust reference implementation for Humans and long-lived AIActors. We argue that any subject that must persist under its own identity and remain independently attributable should have an ActorIdentity that is not replaced by an Account, Credential, Client, AuthSession, IdentityBinding, or runtime instance. SoulAuth therefore treats Humans and long-lived AIActors as first-class identity subjects while keeping authentication distinct from downstream authority. Methodologically, we use a Philosophical Engineering approach that translates conceptual analysis of subjecthood into identity objects, invariants, lifecycle semantics, system responsibilities, implementation boundaries, and inspectable conformance evidence. Evaluation against the fixed SoulAuth v0.1.0 artifact shows that the implementation realizes core boundaries including Human/AIActor first-class identity status, Client/Actor separation, and Authentication/Authority separation, while gaps remain in unified Credential modeling and historical attribution anchored to ActorIdentity. We therefore report partial, not full, architecture conformance.
Read the full abstract
Computational Engineering
Node-Shift-Encoding Genetic Algorithm with fuzzy-enhanced reference tour to solve the bi-objective service-oriented TSP
The Travelling Salesman Problem (TSP) remains a key area of research in combinatorial optimization, with applications in logistics, manufacturing, and service delivery. This paper addresses a bi-objective service-oriented TSP in which the clients' ranks in the delivery path matter. Unlike conventional depot-based TSP formulations, the considered problem does not assume a distinguished depot or a fixed tour origin. To address this setting, we adapt the Miller–Tucker–Zemlin (MTZ)-based formulation and derive an original linearization of the resulting model, enabling its solution with off-the-shelf integer linear programming solvers. This adaptation avoids the rigid tour origin imposed by the conventional MTZ formulation, for which fixing the starting node does not affect the tour cost but can affect the objective in a customer-rank-sensitive TSP. To solve this problem, we present a Node-Shift-Encoding (NSE)-based Genetic Algorithm augmented with fuzzy reasoning to update the reference tour throughout the evolutionary process. Experimental evaluation on TSPLIB benchmarks demonstrates that the proposed method achieves improved performance compared with the classical NSE approach.
Read the full abstract
Probability
Asymptotics for Bivariate Risk Models with Random-Size Clusters of Dependent Claims
This paper investigates finite-horizon tail and ruin asymptotics for models with random-size clusters of dependent main and delayed claims under Levy investments. Under long-tailed, dominatedly varying claim distributions, a conditional, uniformly weighted Ko-Tang condition within clusters, and moment conditions, univariate asymptotics are derived for discounted aggregate claims and ruin probabilities. For two business lines sharing renewal arrivals but having independent claim marks and Levy processes, marginal and joint tail asymptotics are established as thresholds diverge, without restricting their ratio. Results provide asymptotic formulas for ruin in both lines, simultaneous ruin, and ruin in at least one line.
Read the full abstract
Signal Processing
Rethinking Radiomap Blind Prediction with Limited Environment and Configuration Representations
Radiomap blind prediction infers radiomaps from observable representations of the propagation environment and base station (BS) configuration without field measurements. These representations are inherently incomplete and cannot uniquely determine the target radiomap. Under squared loss, we identify the conditional-mean radiomap as the population-optimal deterministic target and decompose domain risk into target-approximation error and irreducible uncertainty. The train-test risk gap motivates propagation priors as cross-domain guidance, although their partial or simplified forms may bias the attainable predictor. We therefore propose RadioDecomp, which treats a prior-guided predictor as a correctable base and uses deterministic residual refinement to learn its remaining predictable discrepancy. We instantiate RadioDecomp as RadioLSR (LoS-Shadow-Residual). Experiments under cross-configuration and cross-environment settings show that RadioLSR is especially effective for cross-configuration generalization and provides overall gains over a controlled monolithic counterpart under cross-environment generalization.
Read the full abstract
High Energy Physics (Phenomenology)
Relocating the SIMP Miracle in the Axion Portal
In the strongly interacting massive particle (SIMP) scenario, dark matter is a pseudo-Nambu–Goldstone boson whose abundance is set by a three-to-two Wess–Zumino–Witten process. We study this scenario in an axion portal and find that the kinetic contact the canonical mechanism assumes is excluded by bounds on sub-GeV axion-like particles (ALPs). The dark sector then freezes out at its own temperature, so the relic abundance no longer fixes the self-interaction cross section but predicts the portal coupling instead. Moreover, the portal operator is Hermitian and even in the ALP field, so no trilinear ALP–pion coupling arises. The contact term then drives $\pi\pi\to aa$ with nothing to cancel against it, and we find the conversion four orders of magnitude faster than a trilinear estimate gives. This disfavours the minimal realisation in which the dark condensate alone generates the ALP mass. The realisation that survives instead makes the ALP slightly heavier than the dark pion. Matching the observed abundance then fixes the flavon vacuum expectation value at $V_\phi\simeq 1.1\times10^{10} \ {\rm GeV}$ and leaves the dark scale open over more than an order of magnitude. The decay $K^+\to\pi^+a$ requires the flavon to charge the leptons alone. The model then predicts a dark matter self-interaction of $0.20\,{\rm cm^2/g}$ at a dark pion mass of $140\ {\rm MeV}$, in a window running from $82$ to $169 \ {\rm MeV}$. The predictive power of the SIMP framework is therefore not lost but relocated, from the self-interaction to the flavon scale.
Read the full abstract
Machine Learning
MUtE: A Dual Framework for Concept Erasure and Counterfactual Interventions
Erasing concept-specific information from representations has been proven useful for mitigating bias or interpreting model decisions. The joint objective is to transform the original representations such that the target concept becomes unpredictable, while maximally preserving concept-unrelated information. In this work, we revisit the optimal bounds of concept erasure to derive a novel class of erasure functions that naturally induce a deterministic, dual counterfactual mapping. Bridging the gap between theoretical optimality and practical representation learning, we design an implementation that imposes a translational bias on counterfactual trajectories - a constraint that aligns with how many concepts geometrically manifest in modern language models. Our framework enables seamless navigation between concept erasure and counterfactual generation. We empirically demonstrate its efficacy in improving downstream algorithmic fairness and generating counterfactual texts.
Read the full abstract
Audio and Speech Processing
AudioICL-Bench: A Benchmark for Large Audio Language Model In-Context Learning
In-context learning (ICL) promises training-free adaptation for audio, where labeling every new condition is costly. Yet existing audio ICL studies largely measure Task Recognition, where demonstrations merely cue pre-trained capabilities, rather than Task Learning, where a genuinely new input-label mapping must be inferred from demonstrations alone. We introduce AudioICL-Bench, a diagnostic benchmark whose per-episode rules are resampled so that no correct answer is recoverable from prior knowledge. Its nine tasks are organized along two axes that separate what must be learned from demonstrations from what must be perceived in the signal, enabling failures to be attributed to either source. Across five Large Audio Language Models, the strongest models readily bind arbitrary sounds to new labels when perception is easy, but collapse on temporal measurement and composing multiple induced rules, revealing two primary capability boundaries: temporal perception and multi-rule composition.
Read the full abstract
Cryptography and Security
A BUD You Can Count On: Efficient, Verifiable Membership Proofs
A block typically changes only a small fraction of blockchain state, yet producing a state root makes validators update an authenticated tree spanning it all. We show how to support verifiable reads from the writes alone. We introduce Block Update Digests (BUDs), that commit to a block's writes and link each key touched in that block to its previous modification. An optional prior-value record lets one successor BUD authenticate the value between modifications, replacing the baseline's two BUD openings. SuperBUDs summarize those writes over longer intervals, letting a reader verify that a value remained unchanged without checking every intervening block. Thanks to BUDs and SuperBUDs, validators can just keep ordinary flat state while still serving proofs of current and historical values, including deletion. We prove soundness and completeness for reads at or after a key's first logged modification when the required proof data is available. In experiments with the baseline format, a 50x increase in state size raises the measured base-BUD cost by just 1.24x. In paired QMDB experiments, lowering its garbage-collection activation threshold raises process CPU time by 1.3-2.7x and device bytes written by 1.6-2.4x relative to its dormant configuration. BUD authentication requires no corresponding relocation of live entries. In the evaluated baseline, two-digest read-layer payloads stay below 800 bytes, and warm hash-path verification takes at most 146 {\mu}s at p99. Authenticated reads need not make validators maintain an authenticated tree over the entire state.
Read the full abstract
Computational Physics
A Geometrically Parameterized Quasi-Stationary 3D Model for High-Frequency Induction Tube Welding
A three-dimensional multiphysics finite element model for the simulation of high-frequency induction welding of tubes is presented. The model couples a time-harmonic $h-\phi$ formulation with a stabilized quasi-stationary advection-diffusion heat transport equation, enabling accurate prediction of electromagnetic and thermal fields under industrial operating conditions. The model incorporates parameterized geometry generation and semi-automated, physics-tailored mesh construction and is implemented using the open-source tools GetDP and Gmsh. Validation against measurements from a commercial induction welding line for AISI 304 stainless steel tubes demonstrates good agreement with operating data. The validated model is subsequently applied to investigate the influence of impeder material by comparing a conventional FeNiZnV ferrite with the soft magnetic composite Ferrotron 559H for the induction welding of AISI 304 stainless steel tubes.
Read the full abstract
Combinatorics
Proper conflict-free choosability of sparse graphs with girth at least seven
A proper conflict-free coloring of a graph is a proper vertex coloring in which every non-isolated vertex has a color appearing exactly once in its neighborhood. A graph $G$ is proper conflict-free $(\mathrm{degree}+2)$-choosable if every list assignment $L$ with $|L(v)|\ge d_G(v)+2$ for each $v\in V(G)$ admits such a coloring from the lists. We prove that every graph with girth at least $7$ and maximum average degree less than $8/3$ is proper conflict-free $(\mathrm{degree}+2)$-choosable. Consequently, every planar graph of girth at least $8$ has this property, improving the previously established sufficient girth bound of $9$.
Read the full abstract
Quantum Physics
Generative Replay Mitigates Sample Starvation in Quantum Architecture Search
Reinforcement learning (RL) can automate quantum architecture search, but its scalability is limited when useful circuit trajectories become rare in the rapidly expanding search space. Existing replay mechanisms reuse observed transitions; the proposed learned model produces additional predicted one step transitions from real state-action seeds. Here we introduce GenQAS, a tensor network-guided RL framework that combines a fixed matrix product state warm-start with prioritized generative replay. A learned local transition model generates synthetic circuit transitions on demand and mixes them with real experience during Double Deep Q-Network updates. Under a random exploration analysis, near ground state circuits occupy a rapidly shrinking region of the accessible state space. We investigate whether real data anchored synthetic replay can improve the effective training signal in this regime. Across chemical Hamiltonian benchmarks from 6 to 12 qubits, GenQAS improves fixed-budget success probability and identifies compact circuits at competitive energy error. At 12 qubits, it improves final success probability by up to $7.0\times$ over passive replay. On a 15-qubit transverse field Ising model, GenQAS increases success probability from $12\%$ to $21\%$. In a noisy 6-qubit BeH$_2$ transfer experiment, generative replay reduces the steps to chemical accuracy by $92.7\%$. These results show that generative replay can mitigate sample starvation in quantum architecture search and support more resource efficient circuit discovery.
Read the full abstract
Computation and Language
The Illusion of Balanced Multimodal Sentiment Analysis: Beyond the Limits of Optimization-Based Methods
Multimodal Sentiment Analysis (MSA) remains constrained by modality imbalance, yet the field continues to rely on optimization-based balancing methods that promise more than they deliver. We provide three contributions: 1) a unified evaluation framework testing gradient and loss-based balancing strategies under controlled settings; 2) a theoretical diagnosis explaining why these methods fail, as they conflate fitting speed with discriminative contribution; and 3) a research agenda toward held-out discriminative modality valuation. Experiments on CMU-MOSI and CMU-MOSEI reveal three shortcomings: no strategy reliably outperforms Late Concatenation; performance is sensitive to hyperparameters; and even ratio calibration fails to yield consistent gains. The core issue is fundamental: loss is not utility, and gradients are not importance. Modality imbalance remains unresolved, motivating utility estimation from held-out performance.
Read the full abstract
Computation and Language
Assessing the Reusability of Public Speech Resources for Low-Resource Languages: A Central Kurdish Case Study
Kurdish is spoken by millions of people, but little technology can read it aloud. A recent study released three Kurdish voices, 35 hours of recorded speech, and a paper describing the work, all free to download. This review checks how well those public files match the paper. The research is careful about its limits, but the files contain several problems: a settings file lists equipment that was never used, test recordings are left unlabeled among training data, and a coding fault mishandles long numbers. The download page also claims a stronger result than the paper reports and recommends one voice for general use. That recommendation matters because Kurdish has major regional and written variation, while these voices were built from three people reading prepared texts. The process therefore removes much everyday and regional speech. English and German benefit from long traditions of dictionaries and linguistic description that help identify wrong pronunciations; Kurdish has far less such support, so software choices can go unchecked. The voices sound fluent, but they represent the reading styles of their speakers rather than Kurdish as a whole. Most of these issues can be fixed using information the team already has, without changing the reported results. Better records would mainly make the work easier for others, especially community linguists, to check and reuse. The license is the main exception: whether audiobook owners allow corrected versions to be shared will affect whether future Kurdish voices can build on this work or must start again.
Read the full abstract
Statistical Methodology
Global and local indicators of spatial connectivity for areal data
Methods for areal data commonly describe spatial structure through global autocorrelation measures or local indicators based on neighboring values. In many applications, however, it is also relevant to know whether high- or low-valued areas form connected spatial structures. Motivated by topological data analysis, we develop global and local indicators of spatial connectivity for observations defined on an areal adjacency graph. We construct thresholded spatial graphs and use the Betti-0 curve to record their number of connected components. We further derive a local decomposition in which each area's contribution to the Betti-0 curve is expressed as an activation term minus a merging term. Each component fusion contributes to the merging term through a symmetric allocation between the entering area and the neighboring areas through which the fusion occurs. Integrating the merging contributions yields area-level indicators of participation in the connectivity of high- or low-valued regions. Global and local inference is based on random relabeling, with a global test for departures from spatial exchangeability and a conditional test for the local indicators. We illustrate the framework using COVID-19 incidence in Italian provinces during two epidemic waves. Significant departures toward greater connectivity were found for both high- and low-incidence areas, with the largest discrepancy corresponding to high incidence during the first wave. The proposed indicators complement conventional measures of spatial association and are applicable whenever the spatial coherence of threshold exceedances or deficits is of interest.
Read the full abstract
Computation and Language
OmniHallu: Unified Hallucination Detection for Cross-Modal Comprehension and Generation in Multimodal Large Language Models
While Multimodal Large Language Models (MLLMs) have achieved remarkable progress across diverse tasks, they suffer from hallucinations where generated outputs contradict or misrepresent input semantics. Existing research typically addresses hallucination detection within a single modality or task type, limiting generalizability. We introduce OmniHallu, a unified hallucination detection framework spanning both comprehension and generation tasks across image, video, and audio modalities. We contribute OmniHallu-Bench, a 10,000-sample benchmark with claim-level human annotations covering six cross-modal tasks: image-to-text (I2T), video-to-text (V2T), audio-to-text (A2T), text-to-image (T2I), text-to-video (T2V), and text-to-audio (T2A). Our multi-agent architecture decomposes model outputs into atomic claims, verifies them through modality-specific experts, and aggregates evidence via structured reasoning. We further propose a preference-optimized trainable verifier that approximates the multi-agent decision boundary, reducing expert calls by 66% with minimal performance loss. Extensive experiments reveal a consistent modality-dependent performance gradient and provide fine-grained insights into cross-modal hallucination patterns.
Read the full abstract
Artificial Intelligence
Sci-MMR: Benchmarking Multi-Step Evidence-Grounded Scientific Reasoning in Multimodal Agents
Autonomous research agents are increasingly expected to search the literature, analyze experimental evidence, and generate scientific hypotheses. These capabilities require multi-step evidence grounded reasoning that progressively acquires, integrates, and verifies evidence before reaching a conclusion. Existing multimodal benchmarks, however, largely evaluate final-answer accuracy, leaving open whether predictions are actually supported by traceable scientific evidence. We introduce Sci-MMR, a benchmark for multi-step evidence-grounded scientific reasoning built on structured argument graphs linking scientific claims, citation-grounded knowledge, visual evidence, and supporting regions. Sci-MMR comprises 235 multi-hop reasoning tasks spanning four scientific disciplines, with an average of nine figure panels per task. Evaluating eight frontier multimodal models, we find that answer accuracy consistently exceeds complete-evidence recovery rate by more than 20%, revealing a substantial gap that answer-only evaluation is structurally unable to capture. Through controlled interventions, we identify two fundamental bottlenecks. First, evidence acquisition: models struggle to extract complete structured evidence from scientific figures, accounting for 57.2% of failures. While cropping tools yield modest gains (+4.5 points), providing gold evidence improves accuracy by up to 37.0 points, indicating difficulty in assembling complete multi-region evidence. Second, evidence integration: models struggle to translate available evidence into correct conclusions, accounting for 31.8% of failures, while even with gold evidence the strongest model achieves only 69.1% accuracy on the hardest tasks. These findings indicate that current answer-centric benchmarks substantially overestimate the evidence-grounded reasoning capabilities of multimodal research agents
Read the full abstract
Computer Vision
From Evaluation to Enhancement: Benchmarking and Improving Think-with-Video Reasoning for Video Generative Models
Video generation has advanced to produce visually compelling and temporally coherent results. Yet, whether these models can genuinely think with video–executing symbolic rules, respecting physical laws, and pursuing intentional goals–remains an open question. Existing benchmarks only partially address this, often conflating visual quality with cognitive correctness. We introduce VWG-Bench (Video World Generalist Benchmark), a comprehensive benchmark spanning 9 reasoning dimensions and 38 fine-grained tasks. To enable precise diagnosis, we design a three-level VLM-as-Judge protocol that independently assesses video-level fluency, task-level rule adherence, and sample-level goal realization. Evaluations of leading models reveal a striking gap: while models achieve strong rendering scores, they consistently fail on logic-heavy and rule-constrained tasks. To address this, we propose Vid-PRE (Video Prompt Reasoner and Enhancer), a model-agnostic prompt rewriter that offloads the cognitive burden of reasoning to a dedicated VLM. Trained via reinforcement learning with purely text-based rewards, Vid-PRE produces concise, constraint-aware prompts without the instability of video-level reward signals. Experiments show that Vid-PRE yields substantial reasoning improvements across multiple generators without architectural modifications. Together, VWG-Bench and Vid-PRE offer a rigorous diagnostic lens and a scalable path toward true think-with-video capabilities. All data and code are publicly available at https://huggingface.co/datasets/KlingTeam/VWG-Bench.
Read the full abstract
High Energy Physics (Phenomenology)
Higgsino dark matter in the Starobinsky supergravity with the MSSM in light of the LUX-ZEPLIN event
We realize a nearly pure higgsino dark matter candidate with mass of about 1 TeV in the Minimal Supersymmetric Standard Model coupled to the Starobinsky supergravity. The recent LUX-ZEPLIN 248 keV nuclear-recoil event has renewed interest in higgsino dark matter. In this framework, the particle spectrum, including the Higgs boson mass, the higgsino mass and its splitting, is connected to cosmic inflation observables via gravitational mediation of supersymmetry breaking and subsequent renormalization group evolution. We show that the Higgs boson mass predicted by the model agrees with the measured value within the quoted uncertainties for a broad range of bino-induced mass splittings. Wino-induced splitting leads to a Higgs mass several GeV higher and is therefore excluded.
Read the full abstract
Computer Vision
Fast and Accurate Monomodal 3D High Resolution Deep Registration of Drosophila Larval Brain Volumes
The larval stage of Drosophila melanogaster is a compact model system for neuroscience whose genetic toolkit allows fluorescent markers to be expressed in defined neural populations, and comparing the resulting expression patterns across animals requires every brain to be registered into a shared anatomical reference space. Existing pipelines for this task are predominantly based on classical registration methods, which perform a new optimization for each volume, often require per-case parameter tuning, and can take minutes per brain, limiting their use as a routine preprocessing step. We present a trained deep registration pipeline that deformably aligns a larval brain to a reference template in a single forward pass at high spatial resolution, on volumes that hold several times more voxels than those learned 3D registration is normally reported on, together with the preprocessing and anatomy-anchored evaluation pipeline required to apply it. Against eleven classical and seven further learned baselines on a held-out collection acquired with different acquisition and quality strata, the proposed pipeline is the most accurate, improving on the strongest classical baseline by 23 percentage points of anatomical landmark-local mutual information. It registers a volume one to two orders of magnitude faster than the classical deformable pipelines, and it retains more of its accuracy than any other method as acquisition quality degrades. The network, its trained weights and the full pipeline are released as the open-source deep larval brain registration framework: https://github.com/agentdr1/deep-larval-brain-reg