Peer Revenue Publish or profit

Papers

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

All of arXiv

Saturday 5 September

Computer Vision

From Gaze to Meaning: A Training-Free AI Agent for Unified Grounding and Explanation

Understanding human attention is fundamental for scene interpretation, yet existing approaches often rely on heavily trained models that lack interpretability. Prior methods struggle to jointly reason about gaze targets, attended objects, and visual grounding without extensive supervision. To the best of our knowledge, this work introduces the first training-free Gaze Target Agent (GTA) for gaze-guided reasoning across tasks such as gaze target prediction, attention localization, and object identification. This is achieved by leveraging pretrained vision-language models, augmenting them with visually guided prompts, and employing a memory-based retrieval strategy for high-uncertainty samples to improve performance without additional training. We evaluate our approach using both quantitative metrics and qualitative results. Quantitatively, our method achieves state of the art performance on the GazeFollow and GazeHOI benchmarks. Qualitatively, our agent provides detailed semantic predictions, predicts the correct targets even when ground truth labels are wrong, and remains flexible without vocabulary constraints.
Read the full abstract

All authors Shayan Nasiriboukani, Sara Atito, Mohammad Nezamipour, Muhammad Awais

Computer Vision

PhysWeep: Does a Video Generator Realize the Physics You Ask For?

Image-to-video generators are often credited with absorbing physical dynamics as implicit world models, a claim the community currently checks with plausibility scores that ask whether a clip looks consistent with real-world motion. Plausibility is the wrong test on its own, because a clip can look natural while encoding the wrong value of the governing physical parameter, and no existing benchmark measures this gap directly. PhysWeep closes it with a fixed, label-free audit, treating a frozen generator as a black box, recovering the realized parameter from generated pixels, and reporting how often generation is trackable at all, how far the realized value sits from the requested one, and which, if either, of the literature's two proposed failure mechanisms the data support. A deterministic-simulator positive control confirms every score is exactly checkable. Applied to three open generators across six sweep axes, PhysWeep finds a specific, reproducible, previously undocumented failure. Conditional on producing trackable motion, two of the three generate confident, well-fit dynamics that converge to one of a small number of fixed, wrong values selected by the sampling seed rather than by the request, reproducing across two independent model families, two physical systems, and an independent tracker. It matches neither the prior reversion nor the case-based clamping the literature anticipates, because the reversion target is seed-conditional rather than a single global default, and a leave-one-out selection rule rejects both; the in-range faithfulness slope is statistically indistinguishable from zero wherever a response is estimable at all. A benchmark averaging over seeds would never see this: each sample is confidently locked to a wrong constant, exactly the failure a plausibility score is structurally blind to. We release the protocol, suite, and analysis code.
Read the full abstract

Systems and Control

Non-parametric Formal Synthesis of Unknown Stochastic Systems: Asymptotic Convergence Guarantees

Data-driven techniques have shown promising potential for checking behavior of complex systems operating in safety-critical domains against safety and other temporal requirements. This paper studies a class of data-driven techniques that are based on learning a representation of the system from data using non-parametric estimation. The proposed approach is able to formally verify discrete-time stochastic dynamical systems against temporal logic specifications only using observation samples and without the knowledge of the model, and provides a probabilistic guarantee on the satisfaction of the specification. We first consider finite abstract representations of the system in the form of Markov decision processes (MDPs) and derive asymptotic convergence guarantees between the transition probabilities of the abstract MDP and their estimation using Bernstein's inequality and statistical properties of non-parametric estimators. We then propose theoretical results for estimating the asymptotic upper bound of the Lipschitz constant (LC) of the stochastic system, which can determine the size of the finite abstract MDP for a given precision error. Under appropriate assumptions, our results prove that the asymptotic convergence rate of the estimations is $O(n^{-1/(3+d)})$ for both the transition probabilities and the LC, where $\mathsf d$ is the dimension of the system and $n$ is the data scale. By integrating these results, we can guarantee the asymptotic closeness in formal verification and policy synthesis performed on the original system and its finite abstraction based on the size of the dataset. Multiple case studies are presented to validate the effectiveness of the proposed method.
Read the full abstract

Analysis of PDEs

Logarithmic capacity and torsion in planar shape optimization

We study a family of functionals involving a product of logarithmic capacity and torsional rigidity. Using several geometric inequalities, we identify the discs as the only maximizers among planar convex sets for certain values of the parameters. Finally, we prove regularity of maximizers in the non-critical case. This is done by establishing a shape-differentiability theorem for the logarithmic capacity in the class of planar convex bodies.
Read the full abstract

Networking

Rethinking Channel Charting: A Graph Perspective

Channel charting is a self-supervised framework that learns low-dimensional spatial representations from high-dimensional channel state information. We revisit channel charting from a graph-theoretic perspective, and show that the position-diffusion objective is equivalent to a graph Laplacian smoothness functional. We propose a Graph Neural Network (GNN) formulation that replaces the Siamese network's global geodesic dissimilarity objective with a graph smoothness objective. We consider the reformulated objective to be the position diffusion objective. Without diverging from the original optimization objective, the GNN replaces the quadratic-cost self-correlation encoding with linear-cost message passing over an Angle-Delay Profile (ADP)-similarity graph, achieving comparable positioning accuracy with 512 times fewer parameters. Using Laplacian spectral analysis, we demonstrate that obstacles compress the ADP graph spectrum, while the GNN acts as a spectral decompressor against obstacles and other environmental semantics, but as a compressor against excessive ADP embedding space. Beyond this, the second eigenvector of the learned embedding encodes the line-of-sight/non-line-of-sight boundary rather than spatial coordinates.
Read the full abstract

All authors Yifei Jin, Yuxin Zhao, Dandan Hao, Gabor Fodor

High Energy Physics (Phenomenology)

Planck-Scale Signatures in Vacuum Neutrino Oscillations

We present a closed-form perturbative framework for two-flavour vacuum neutrino oscillations in the presence of a weak, flavour-blind Planck-scale mass correction generated by the dimension-five Weinberg operator of the Standard Model Effective Field Theory. The perturbative treatment is developed in direct structural analogy with approaches used to describe weak Earth-matter effects. We derive simple analytical expressions for the modification of the oscillation length, the oscillation phase, and the flavour-conversion probability, valid for arbitrary neutrino mixing and baseline lengths. Unlike matter-induced perturbations, the Planck-scale correction is constant along the neutrino trajectory. Consequently, its effect on the transition probability remains bounded and independent of the baseline at the amplitude level, while the associated modification of the effective mass-squared splitting produces a phase mismatch that increases linearly with propagation distance. We further present illustrative order-of-magnitude sensitivity projections for next-generation long-baseline and reactor neutrino experiments, including DUNE, Hyper-Kamiokande, and JUNO. Our results indicate that the accumulated phase drift provides a particularly promising observable for probing Planck-scale mass corrections and for distinguishing a possible quantum-gravity contribution from conventional matter-induced effects.
Read the full abstract

Optimization and Control

Weak Convexity and Proximal Bundle Methods for Nonsmooth Policy Optimization in Robust Control

We study policy optimization for discrete-time robust $\mathcal{H}_\infty$ control with static output-feedback, and present the first feasibility-preserving algorithm with a deterministic, non-asymptotic complexity guarantee. This problem naturally leads to a nonsmooth and nonconvex optimization over the set of stabilizing feedback gains. We first establish several structural properties of the $\mathcal{H}_\infty$ cost. In particular, we show that the cost is weakly convex on every convex subset of a sublevel set. For the state-feedback case, we further establish a weak Polyak–Łojasiewicz inequality, which ensures that every stationary point is globally optimal. Building on these properties, we develop a proximal bundle method for $\mathcal{H}_\infty$ policy optimization. The proposed method can be viewed as an implementable approximation of the proximal point method and uses only function value and subgradient information. We show that all iterates remain stabilizing and establish a deterministic non-asymptotic complexity bound of $\mathcal{O}(\max\{\eta^{-4},\epsilon^{-2}\})$ for finding an $(\eta,\epsilon)$-stationary point. Numerical experiments illustrate our theoretical results.
Read the full abstract

High Energy Astrophysical Phenomena

The First Multi-Messenger Nova: External Shocks, TeV Photons, and Neutrinos in the Next Eruption of T CrB

T Coronae Borealis is the most compelling anticipated Galactic recurrent nova and one of the best opportunities to test whether nova eruptions can become genuine multi-messenger transients. Recent observational work on classical novae has shown that the GeV $\gamma$-ray luminosity correlates strongly with the differential velocity between interacting outflows, suggesting that shock power depends strongly on velocity contrast. We apply this empirical framework to T CrB and compare it with the 2021 eruption of RS Ophiuchi, the only nova securely detected at TeV energies. We argue that T CrB may lie at the extreme high-$\Delta v$ end of the nova population and that its strongest shock should arise at the external interaction between the nova ejecta and the slow circumbinary medium. This makes the external shock the most likely site of any TeV emission, while the GeV signal may contain contributions from both internal, if present, and external shocks. Because T CrB is substantially closer than RS Oph, it is an especially favorable target for TeV detection. The neutrino case is promising but more uncertain because it depends sensitively on the density and structure of the circumbinary material. We also revisit the recurrence time of T CrB. Using the observed scatter of better-sampled recurrent novae, we infer a representative population-scale recurrence of $T_{\rm rec}\simeq80\pm12$ yr. Taken together, its proximity and potentially extreme shock velocities make T CrB perhaps the best opportunity for a nova to become the first securely established multi-messenger nova.
Read the full abstract

All authors E. Aydi, P. Craig, K. V. Sokolovsky, L. Chomiuk, C. C. Cheung, L. Izzo, J. D. Linford, B. D. Metzger, S. Mohamed, I. Molina, K. Mukai, K. J. Shen, J. L. Sokoloski

Dynamical Systems

Independence of multipliers via degenerations of rational maps

Using degenerations of rational maps and local asymptotics of periodic multipliers, we prove that for every $d\ge 2$, the multipliers of any $2d-2$ distinct periodic orbits of degree $d$ rational maps are algebraically independent over $\mathbb C$, provided that at most $d$ of the selected orbits are fixed points. This condition is sharp because of the Holomorphic Index Formula that relates the multipliers of the $d+1$ fixed points. The result of this paper removes the additional restrictions on periods present in earlier work. The proof proceeds by induction on the degree, using a one-hole degeneration for the induction step and a three-hole degeneration in an exceptional degree four case.
Read the full abstract

Mathematical Software

Gradient Reconstruction in Lattice Boltzmann Methods for Systems of Conservation Laws

The automatic derivation turns a declared system of conservation laws into a lattice Boltzmann scheme, giving each conserved physical quantity a set of q populations whose linear equilibrium embeds the physical flux in their first moment. When the flux depends on gradients of the conserved state, those gradients are supplied by tracking them as additional transported fields. Since lattice Boltzmann is commonly memory-bound, these additional degrees of freedom reduce the achievable throughput. To reclaim it, we reconstruct the gradients from the moment structure of the equilibrium instead of transporting them. To leading order, the first moment of a conserved quantity's non-equilibrium part carries its gradient. Because the reconstructed flux enters its own equilibrium reference, that moment is the image of the gradient under a linear operator built from the diffusive-flux Jacobian. The reconstruction is that operator's algebraic inverse, generic across gradient-form constitutive closures and generated automatically from each declared flux. The resulting scheme carries the conserved quantities alone, reconstructing the required gradients from the populations and forming the fluxes locally. It converges at second order in double precision across advection-diffusion-reaction, Allen-Cahn, Navier-Stokes, resistive magnetohydrodynamics and homogenized compressible Navier-Stokes-Fourier systems, matching the accuracy of gradient tracking at equal resolution. On an NVIDIA RTX A5000 it is up to 3.7 times faster in single precision, the memory-bound kernels reaching up to 97% of peak memory bandwidth.
Read the full abstract

Materials Science

Zonal dislocations in Laves phases: A coupled synchro-shear slip mechanism

Synchro-shear is the primary plastic deformation mechanism in Laves phases at elevated temperatures, mediated by synchro-Shockley partial dislocations-zonal dislocations that proceed via localized events such as kink-pair nucleation and propagation. Using atomistic simulations, we identified a novel slip mechanism in Laves phases, namely coupled synchro-shear slip, involving the synchronized glide of two synchro-Shockley partial dislocations on adjacent slip planes, leading to the formation of extrinsic stacking faults. High-resolution scanning transmission electron microscopy revealed the extended core structures consistent with coupled synchro-Shockley partial dislocations bounded by extrinsic stacking faults in C15 NbCr2 and their involvement in twinning. These results highlight the critical role of coupled synchro-shear slip in enabling phase transformations between Laves polytypes and in governing twinning behavior, providing new atomistic insight into the kinetic nature of plasticity in topologically close-packed intermetallic phases.
Read the full abstract

All authors Sang-Hyeok Lee, Mariano Forti, Thomas Hammerschmidt, Gang Liu, Julien Guénolé, Siyuan Zhang, Gerhard Dehm, Sandra Korte-Kerzel, Zhuocheng Xie

Optimization and Control

Algorithmic Optimality Guarantees for Nonsmooth $H_\infty$ Output-Feedback Policy Search

We study continuous-time full-order dynamic output-feedback $H_\infty$ policy search, a nonconvex and nonsmooth problem. Direct policy search is a central paradigm in reinforcement learning and continuous control, but rigorous guarantees remain scarce in robust output-feedback settings. The $H_\infty$ problem is a canonical benchmark because it captures disturbance attenuation and robustness while exposing the hard nonsmooth geometry of policy-space optimization. We prove that on the exact identity-gauge slice of the extended convex lift, $\varepsilon$-stationarity yields $O(\varepsilon)$-suboptimality on compact exact slices, which in turn yields convergence-rate guarantees for nonsmooth policy-search methods. This result addresses the finite-time optimality-gap question raised by Guo and Hu [2022] in the more general dynamic output-feedback $H_\infty$ policy-search setting. We further use the established value equivalence supplied by extended convex lifting to formulate a nonstrict-feasibility bisection method with one final strict-feasibility recovery step, yielding an explicit $\varepsilon$-optimal stabilizing controller. These results provide a quantitative and algorithmic strengthening of prior qualitative optimality theory for nonsmooth $H_\infty$ policy search.
Read the full abstract

Optimization and Control

Fast PAC Global Optimization via Restarted Langevin: Exploration, Exploitation, and Degenerate Cooling

This paper addresses a fundamental question in non-convex optimization: How should a stochastic optimizer allocate computation between global exploration and local exploitation? We study this question in continuous time, using Langevin dynamics for global exploration and deterministic gradient flow for local exploitation. The simplest algorithm in this class is the best-state Langevin–gradient method: several Langevin trajectories explore the objective landscape, the best state encountered is retained, and a single gradient trajectory then refines this state to high terminal accuracy. Our main objective is to characterize the computational work required to achieve a prescribed accuracy with prescribed confidence. We develop low-temperature approximations for this probably approximately correct (PAC) work–accuracy tradeoff, both for the best-state method and for related Langevin schemes. Global exploration is governed by energy barriers, through spectral-gap and Eyring–Kramers asymptotics. The resulting approximations expose the competing effects of temperature, computational budget, and the number of exploratory trajectories, and quantify their dependence on the confidence parameter. In contrast, terminal accuracy is largely separated from global exploration: once a suitable region of attraction has been reached, gradient flow requires only $O(\log(1/\varepsilon))$ additional work to reach accuracy at least $\varepsilon$. Numerical experiments on challenging objectives serve to compare algorithms and to test the quantitative predictions of the asymptotic theory. They identify regimes in which the low-temperature approximations provide useful guidance, as well as regimes exhibiting substantial temperature sensitivity and potentially significant high-dimensional limitations.
Read the full abstract

Machine Learning

EgoNeMo: Transferable Map of Pedestrian Dynamics via Egocentric LiDAR Scan

This paper proposes a transferable Map of Dynamics (MoD) framework that generalizes to unknown environments using only egocentric 3D LiDAR point clouds to overcome the long-standing limitation of traditional MoD methods. While MoDs are essential for encoding human motion characteristics to enable accurate pedestrian trajectory prediction or safe robot navigation, traditional approaches suffer from site-specificity, requiring exhaustive trajectory accumulation at every new location. Extending recent advances in neural implicit modeling, our framework trains a continuous, LiDAR-based MoD estimator across diverse environments. To mitigate the inherent sparsity and temporal bias of real-world trajectory data, we introduce a position-balanced sampling strategy and a multi-task learning architecture that jointly predicts motion distributions and a spatial frequency score map. The latter is further augmented by visibility-aware losses to compensate for incomplete observation data. Comprehensive experiments demonstrate that our method effectively reconstructs underlying motion maps even in unknown locations from a single instantaneous LiDAR scan, despite highly sparse training data. Finally, we show that our improvements enhance the reliability of downstream trajectory prediction.
Read the full abstract

All authors Azusa Sawada, Allan Wang, Hideo Saito, Aaron Steinfeld

Artificial Intelligence

Predicting Wind Turbine Power Using Machine Learning and Weather Forecasts

Offshore wind turbines are widely used to generate renewable energy, but their maintenance can result in decreased efficiency due to forced shutdowns. Accurate wind turbine power predictions can identify periods of low power that would be ideal for scheduling maintenance. However, the effects of data volume, feature selection, and data preprocessing on the performance of such power prediction models have not been thoroughly studied. Besides, current models have limited transferability between different wind turbines. Therefore, this study developed a baseline Linear Regression for performance comparison with a more complex Artificial Neural Network model to predict the power output of a wind turbine, using weather conditions only to enhance applicability. A range of data preprocessing techniques were studied, and models were trained on one month and one year of data to determine the effects of data preprocessing and volume on model performance. Feature selection was explored using a Random Forest Regressor. The best results from the different models showed that the Artificial Neural Network models provided the highest accuracy, with an R2 score of 0.98 and a low Mean Absolute Error of 194, when compared with the baseline model (R2 score of 0.94 and Mean Absolute Error of 441). The model performance is comparable to the range of results in past studies, with the advantage that the proposed method leverages a separate weather dataset from a nearby weather station, enabling future applications for similar wind turbines in different locations. The Artificial Neural Network model was then used to identify 4-h periods of low power predictions over 2 months (simulating application for future periods), providing power output savings of approximately 2000 kW for each maintenance event.
Read the full abstract

All authors Khivishta Boodhoo, Isaac Triguero, Josh Plumbly, Bruce Nicolson, Nicholas Watson

Optics

Cell Reproduction in a Dark Optical Trap

Optical tweezers are a versatile tool in the domain of cytology, enabling trapping and manipulation of individual cells. However, the incidence of laser light causes photodamage to biological matter, even when operating at low optical powers and over short periods of time. Here, we demonstrate stable trapping of single living Saccharomyces cerevisiae yeast cells in vitro using a dark optical trap operating in the repulsive regime of light-matter interactions. In contrast to standard tweezers, our dark optical trap confines cells for hours with negligible disruption to their morphology and reproduction cycle, opening up new possibilities for long duration experiments with living organisms such as the observation of cell reproduction under laser trapping.
Read the full abstract

All authors Ariel Hertz, Gabriel Dias, Carlos L. R. Fragoso, Anna De Falco, A. Z. Khoury, Thiago Guerreiro

Artificial Intelligence

SCIRIGOR:Evaluating Open-Ended Scientific Analysis Beyond Final Scores

Scientific coding agents produce interdependent code, results, figures, and claims, yet evaluating final outputs alone does not establish whether their conclusions are scientifically supported. We formulate evidence-grounded multimodal scientific analysis, requiring agents to produce executable analyses and claims supported by results and visualizations from the same run. We introduce SciRIGOR, an evaluation framework and benchmark comprising 100 cases from scientific articles across six domains and 17 subfields. The framework reconstructs typed evidence graphs, separates artifact fidelity from relational validity, and scores complete claim-support paths while localizing the earliest unsupported relation. Source-grounded alternative paths accommodate scientifically equivalent analyses and visualizations. We evaluate 11 agent/model configurations. On full-benchmark runs, claims agree with faithful and unfaithful results at nearly identical rates (91.8% versus 91.0%). Yet no system exceeds 62.6% on the soft evidence-chain score or 18.0% strict whole-chain success. These findings show that internal coherence does not establish scientific correctness: evaluation must verify support along the complete data-to-claim path.
Read the full abstract

All authors Bowen Liu, Shuo Nie, Bodong Du, Xiaomeng Li

Optics

Optimization Algorithm-Assisted Nanostructure Engineering for Performance Enhancement of Thin-Film Solar Cells

A computation-based study is presented that utilizes the finite-difference time-domain (FDTD) method to perform optical simulations. These simulations were used to observe the effects of core-shell nanostructures with triangular cross-sections embedded inside the light absorbing layer of thin-film solar cells (TFSCs), for improving the optoelectronic performance. Despite extensive research, the relationship between the geometric parameters of light trapping nanostructures and TFSC performance is not yet fully elucidated. To assist in finding the optimal set of parameters, optimization algorithms are used to find a set of parameters that provide maximal short-circuit current density ($J_{SC}$). Three separate algorithms were applied, and the results were scrutinized in-depth with a sensitivity analysis, observing the effects of small changes in the optimal parameters. The optimal parameters of the nanostructure produced $J_{SC}$ values ranging from 28.90 to 35.39 mA/cm$^2$ compared to 13.69 mA/cm$^2$ for a reference TFSC with no nanostructure present. From the results of this study, it can be inferred that optimized periodic grating structures can provide substantial improvement in performance of TFSCs.
Read the full abstract

All authors Abyaz Karim, Sajid Ahmed Chowdhury, A. F. M. Afnan Uzzaman Sheikh, Asif Al Suny, Mustafa Habib Chowdhury

Computer Vision

One Perturbation Is Not Enough: Identifiability and Blind Baselines for Behavioral AI Evaluation

Behavioral evaluations perturb an input and read the induced change in the output in order to certify that a system uses that input. We show that the number of perturbations such a certificate requires is fixed, and that reporting a single perturbation cannot supply it. Where a response ratio is a property of the policy rather than of the test items, the behavioral record is a linear measurement of an exponent vector recording how much the output depends on each input, so perturbations identify input use exactly when their logarithms span the input space. At least $n$ are needed for $n$ inputs, an incomplete design confuses precisely the policies differing along the kernel of its design matrix, and sharpening a perturbation never substitutes for adding an independent one. We also derive in closed form the score such a test awards a policy that reads nothing, which is far from zero and which none of the probes we survey reports. Instantiating this where the correct response is fixed by dimensional analysis, we run a complete identifying set of three perturbations on three vision–language models reporting a physical quantity from video. All three score far below their own blind bound rather than above it, because each defaults to one of a small set of round calibration values that never matches what the scale asserts; none moves its relabeling response by a single exponent, and none is separable from the same model instructed to ignore the video.
Read the full abstract

All authors Rasul Khanbayov, Mariam Sohail, Ahmed Abdala, Hasan Kurban

Artificial Intelligence

Customer Relationship Intelligence: Integrating CRM and MDM for Enhanced Customer Engagement

This study examines how Customer Relationship Management (CRM), Master Data Management (MDM), and Customer Knowledge Management (CKM) jointly constitute a Customer Relationship Intelligence (CRI) framework for enhanced Customer Engagement (CE). A cross-sectional survey of 100 participants across retail, healthcare, IT, and telecommunications sectors was analysed using Spearman rho correlation and ordinal logistic regression (IBM SPSS). Bivariate correlations were weak and non-significant (r<0.19, p>0.06). Regression identified CRM (beta=0.717, p=0.002) and CKM (beta=0.581, p=0.009) as significant positive predictors of CE; MDM showed a positive but non-significant direct effect (beta=0.346, p=0.071). The model explained 20.5% of CE variance (Nagelkerke R^2=0.205). Parallel mediation analysis (Hayes PROCESS Model 4, 5,000 bootstrap samples) found no significant indirect effects of MDM on CE via CRM (IE=0.021, 95% BC CI [-0.072, 0.121]) or CKM (IE=0.032, 95% BC CI [-0.061, 0.126]); Hypothesis H4 was not supported. CRM and CKM emerge as the principal drivers of CE within the CRI framework, while MDM functions as a foundational data quality enabler whose strategic value is realised through its enabling effect on CRM execution and knowledge management. Findings should be treated as exploratory given the sample size and cross-sectional design. Future research should replicate with larger sector-specific samples and longitudinal designs, particularly in regulated BFSI contexts where MDM architecture is shaped by data governance mandates.
Read the full abstract

Artificial Intelligence

MVFA: A Multi-View Text-Guided Multimodal Fusion LLM Adapter for Sentiment Analysis and Emotion Recognition

Multimodal sentiment analysis and emotion recognition in conversations demand effective modeling of heterogeneous interactions across textual, acoustic, and visual modalities. Although large language models (LLMs) offer powerful language understanding, adapting them to multimodal affective computing remains challenging: full-model fine-tuning is computationally prohibitive, while many existing lightweight adapters fail to preserve rich textual cues during cross-modal fusion. To address these limitations, we propose the multi-view text-guided multimodal fusion adapter (MVFA), a parameter-efficient framework that augments frozen LLMs with strong multimodal reasoning capability. MVFA first constructs complementary text views via max pooling, mean pooling, and attention pooling; these views then guide cross-modal interactions with audio and visual features. The fused multimodal representations are subsequently compressed into a compact set of learnable pseudo-tokens through an Enhanced Q-Former Fusion Module. Using ChatGLM3-6B-base as the primary backbone, we further validate MVFA on LLaMA2-7B and Qwen3-8B to examine its portability across multiple frozen LLM backbones. MVFA is evaluated on three challenging datasets: CH-SIMS V2.0, MELD, and CHERMA. Experimental results demonstrate that MVFA achieves state-of-the-art performance on key metrics while updating only a small fraction of parameters. Specifically, it attains 84.62% Acc2 and 84.59% F1 on CH-SIMS V2.0, 67.36% Acc and 66.03% WF1 on MELD, and 74.66% Acc on CHERMA. These findings establish multi-view text-guided fusion as an effective and scalable paradigm for parameter-efficient multimodal LLM adaptation in affective computing. The code is publicly available at https://github.com/Overwhelm1208/MVFA.
Read the full abstract

All authors Pengfei Shao, Jisheng Dang, Jiawen Fang, Ning Liu, Wencan Zhang, Bimei Wang, Jingwen Zhao, Jianhuang Lai, Qi Tian, Tat-Seng Chua

Computer Vision

SolarBench: A global solar energy nowcasting benchmark

As the share of solar power grows, nowcasting weather-driven solar variability becomes critical for reliable energy system operation. State-of-the-art approaches increasingly apply deep learning to sky camera and geostationary satellite observations, but fragmented datasets and inconsistent evaluation make it difficult to determine whether reported improvements generalize across climates, cloud regimes, and photovoltaic (PV) systems. Here we introduce SolarBench, an open global benchmark for image-based solar nowcasting. SolarBench harmonizes more than six million sky and satellite images from 11 diverse sites spanning a decade, together with irradiance or PV output and auxiliary atmospheric data. An accompanying toolbox supports reproducible data access, processing, model development, and evaluation. Using SolarBench, we benchmark representative models and reveal a gap between average forecasting accuracy and the ability to capture rapid solar fluctuations. We further quantify predictability across cloud regimes and demonstrate data-efficient adaptation to new PV systems. SolarBench provides an extensible foundation for fair comparison and methodological innovation in solar nowcasting.
Read the full abstract

All authors Yuhao Nie, Stephen Campbell, Quentin Paletta, Liwenbo Zhang, Tao Jing, Samer Chaaraoui, Jonathan Giezendanner, Andea Scott, Tao Sun, Cong Feng, Max Aragon, Jacques Camier, Adam Jensen, Florian Kotthoff, Yuexing Yang, Yang Ming, Mengying Li, Stefanie Meilinger, Yupeng Wu, Adam Brandt, Sherrie Wang

Machine Learning

Spectral Prioritized Sweeping in Nonstationary Reinforcement Learning

Prioritized Sweeping (PS) accelerates model-based reinforcement learning by selecting backups according to Bellman residual magnitude. In nonstationary reward settings, however, the canonical priority score is shortsighted: after a localized reward shift, residuals propagate only through realized backups, so bottlenecked or topologically distant state estimates may remain static under a limited replanning budget. We introduce the Graph Topology Augmentation framework, which employ the graph's resolvent and its diffusion semantic, to augment the inquired signal. Our application, Graph Topology Augmentation for Prioritized Sweeping (GTA-PS), or which the alias Spectral Prioritized Sweeping (SPS) might be more universal, provides a drop-in ordering score for the setting of fixed dynamics and changing state rewards. GTA-PS uses a smootherized policy, inducing a transition chain, with its in- and out-Laplacian. The standard priority key is augmented with a mixing of regularized Laplacian inverses diffusing the residual magnitude. Furthermore, the topology contribution is annealed by a scheduler based on the Second Largest Eigenvalue Modulus (SLEM), allowing its scale to adapt to the chain's mixing regime. We prove that the forward potential coincides with geometric discounted residual propagation and show that GTA-PS gives active priority instantly to all states. Tabular experiments on FourRooms and GARNET domains demonstrate improved replanning efficiency over standard PS under both exact DP and Dyna-style host planners.
Read the full abstract

General Relativity and Quantum Cosmology

Heat distribution of quantum fields interacting with Unruh-DeWitt detectors

We introduce an operational framework for heat-exchange statistics between a quantum field and an Unruh-DeWitt detector in Minkowski spacetime. Using an interferometric protocol, we define the characteristic function of heat without projective measurements, consistently with relativistic causality. We derive perturbative expressions for this function and the heat distribution. In the vacuum, the statistics coincides with the low-intensity expansion of a unidirectional Poisson law, while for Kubo-Martin-Schwinger (KMS) states it has a bidirectional Poisson structure up to second order in perturbation theory. We study how fluctuation relations emerge from detector properties and field correlations, identifying regimes where they reduce to known heat-exchange fluctuation relations. We derive the entropic Landauer inequality as a specialization of the algebraic entropy-balance theorem of Jakšić and Pillet and verify it numerically within the perturbative detector-field interaction. We compare its tightness with a complementary Landauer bound derived from the full counting statistics of heat, analyzing their dependence on the detector's initial state and interaction time. We also formulate the first and second laws for the explicitly switched detector-field dynamics, identifying the external switching work that reconciles the field-energy gain with the detector-energy change. In the gapless and delta-switching regimes, where the dynamics admits a non-perturbative treatment, the characteristic function coincides with that of a bidirectional Poisson process, yielding an exact fluctuation theorem determined by the KMS condition. These results provide a consistent framework for heat statistics in relativistic quantum field theory and clarify the emergence of fluctuation relations.
Read the full abstract

Cosmology and Nongalactic Astrophysics

Probing Dark matter-Baryon Interaction with S301–the Fastest Known Star in the Milky Way

The recent discovery of S301 provides an unique opportunity to investigate dark matter-baryon interaction in the innermost region of the Milky Way. S301 follows a highly eccentric orbit around Sgr~A*, and its pericenter distance is only around 12 AU with velocity about $25000\,{\rm km/s}$. These extreme properties make S301 highly sensitive to dark matter-baryon scattering in the center of galaxy, where the dark matter density may be strongly enhanced. We have calculated the orbit-averaged energy deposited rate in S301 via dark matter-proton scattering, and show that over a broad dark matter parameter space, cross sections satisfying the existing bounds from terrestrial experiments and astronomical observations can inject sufficient energy to modify the luminosity and evolution of S301. Our results establish S301 as an exceptionally sensitive probe of dark matter-baryon interactions and open a new avenue for probing dark matter interactions in high-density and high-velocity astrophysical environments.
Read the full abstract

Mathematical Physics

Renormalization Group on Wiener Space: Spectral Theory and Universality

We develop a rigorous renormalization group (RG) formalism on Wiener space. We introduce an RG operator $\mathcal R$ acting on probability measures over continuous paths, whose fixed point is the Wiener measure. We linearize $\mathcal{R}$ around the Brownian fixed point and analyze the spectral structure of the resulting operator $\mathscr{L}$. Its eigenvectors are too singular to be realized as honest measures on path space, and we therefore develop a generalized spectral theory within the framework of white noise analysis (Hida calculus). The eigenvectors of $\mathscr{L}$ are realized as Hida distributions, satisfying $\mathscr{L} U=\lambda U$ in the weak sense. This analysis yields structural results such as a spectral gap and a diagonal-concentration property of the eigenvectors. Moreover, we identify a family $\{\mathfrak{U}_n\}_{n \geq 0}$ of eigenvectors with eigenvalues $\lambda_n = 2^{1-n/2}$, consisting of Wick polynomials of white noise formally expressed as $\mathfrak{U}_n = \int_0^1 {:}\dot W(t)^n{:}\, dt$, and rigorously constructed as Hida distributions. We then use this spectral structure to uncover a finer, second layer of universality in the Donsker invariance principle: not only is the convergence of random walks towards the Brownian scaling limit universal, but the entire hierarchy of leading corrections to the Brownian limit is universal as well, governed by the eigenpairs $(\lambda_n, \mathfrak{U}_n)$. We also show, at a formal level, that the framework developed here extends to Gibbs-type perturbations of the Brownian fixed point, in the spirit of self-interacting quantum field theories. In particular, we recover the standard irrelevant/marginal/relevant classification, usually obtained in the physics literature by power-counting, solely from the spectral analysis of $\mathscr{L}$.
Read the full abstract

Machine Learning (Statistics)

Recovering linear images of sparse signals from indirect observations

In this paper, we develop and analyze techniques for recovering a linear image $Bx$ of an unknown signal $x$ from indirect noisy observation $\omega=Ax+\xi$. It is a priori known that $x\in \cX$, a given convex compact set, and that $x$ is $s$-sparse—has at most $s$ nonvanishing entries. The proposed estimates belong to a large family of recovery routines by $\ell_1$-minimization. However, unlike the classical result describing performance of such estimates, we do not make any special (and hard to check) assumptions about the sensing matrix $A$ such as nullspace or Restricted Isometry condition and the like. As a consequence, parameters of the estimates and the upper bounds on their risks are not available in a closed analytic form, but are delivered instead by efficient computation as solutions to explicit convex optimization problems.
Read the full abstract

Distributed Computing

Diamond Agent: Agentic Control of Federated HPC Resources as a Service

Efficiently aggregating and orchestrating computing power across heterogeneous clusters for HPC workflows faces four practical challenges: preserving workflow context across independently administered clusters, moving large datasets between sites, reasoning about site-specific environments and scheduler policies, and exploiting live queue and resource states for efficient task scheduling. To this end, we design Diamond Agent, an agentic system that enables intelligent execution of HPC workflows across heterogeneous clusters with typed skills as the interface. Diamond Agent provides an agent-facing workspace and skills that unify cross-site resource discovery, resource specification, data movement, task execution, and result retrieval. A centralized Diamond Agent instance can operate multiple supercomputers without being deployed separately on each login node. Diamond Agent translates high-level agent actions into valid site-specific executions, moves data through Globus Transfer, and uses live system capability and queue information to select feasible placements. Its event-driven continuation mechanism decouples agent actions from long-running batch jobs: persistent services monitor remote execution and resume the agent only when a result or decision-relevant event is available. We experiment with 27 hours of telemetry and 19 matched multi-site submission rounds comprising 83 jobs across four production supercomputers. Compared with a fixed-site baseline, Diamond Agent reduces the median additional completion time relative to the fastest observed placement from 42 seconds to 4 seconds, a 10.5x reduction.
Read the full abstract

All authors Haotian Xie, Junlin Chen, Mingkai Zheng, Yifan Zhu, Minu Mathew, Max Burnette, Yadu Babuji, Volodymyr Kindratenko, Shivaram Venkataraman, Kyle Chard, Ian Foster, Zhao Zhang

Differential Geometry

A Ballantine-Type Factorization of Diffeomorphisms

We show that on a compact, connected, Riemannian manifold without boundary, every isotopic to the identity diffeomorphism can be written as a finite composition of optimal mass transport maps. This result represents a non-linear infinite dimensional generalization of Ballantine's factorization results.
Read the full abstract

Quantum Physics

Quantum lattice Boltzmann method via density-matrix encoding for fluid simulation with wall boundary conditions

The quantum lattice Boltzmann method (QLBM) holds promise for efficient fluid simulations, yet the treatment of realistic boundary conditions, particularly solid wall boundary with arbitrary geometry, remains a critical open challenge. We propose a QLBM algorithm that incorporates wall boundaries as well as inlet/outlet velocity conditions. Building upon the existing QLBM with ensemble transformations, we adopt a density-matrix encoding that offers greater flexibility in circuit design and enables the implementation of non-unitary operations through Kraus operators. For boundary enforcement, we design a quantum implementation of the half-way bounce-back scheme and introduce a component exchange step prior to the streaming step, which naturally enforces the no-slip condition without modifying the transport operation. Inlet and outlet conditions are imposed via controlled SWAP operations combined with ancilla registers prepared in the prescribed velocity states. Our QLBM is quantitatively validated through simulations of two-dimensional flows past a backward-facing step, a cylinder, and an obstacle with complex geometry, demonstrating both accuracy and extensibility of our boundary algorithm.
Read the full abstract