Peer Revenue Returns subject to peer review

Papers

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

All of arXiv

Wednesday 2 September

Computer Vision

YesTrack: Referring Multi-Object Tracking via MLLM-based Yes/No Verification

Referring multi-object tracking (RMOT) aims to track every instance in a video that matches a given language expression. Despite the recent integration of multimodal large language models (MLLMs) to enhance generalization, existing methods predominantly relegate them to the role of caption generators, necessitating external modules for final decision-making. This paradigm not only introduces extra latency but also severely underutilizes the inherent vision-language alignment capabilities of MLLMs. To address these limitations, we propose YesTrack, a novel two-stage RMOT method that reformulates referring as a discriminative task, directly leveraging MLLMs for Yes/No verification without explicit text generation. To further enhance the reliability and efficiency of this MLLM-based verification, we introduce two lightweight temporal consistency constraints: Temporal Confidence Prior (TCP) and Temporal Reference Propagation (TRP). We further validate the generality of this discriminative paradigm by proposing YesTrack-MOT, a straightforward yet highly effective instantiation for generic multi-object tracking (MOT). Experiments on Refer-KITTI and Refer-KITTI-V2 show that YesTrack significantly outperforms existing state-of-the-art methods while maintaining high efficiency, even when implemented with the smallest variant of Qwen3-VL. Code is released at https://github.com/ggbondrighthere24/YesTrack.
Read the full abstract

All authors Quansheng Hu, Qin Sun, Qiansen Dai, Jin Ding, Wan Zhang, Xue Zhou, Jianxiao Zou

Optimization and Control

Characterizations and Complexity of Minimum Forward and Integer Cycle Bases

The cycle space of a directed graph is generated by a cycle basis, where, in general, cycles are allowed to have both forward and backward arcs. In a forward cycle, all arcs must follow the given direction. Several open questions remain regarding the complexity of the minimum cycle basis problem, in particular the minimum-weight integral cycle basis problem, and the minimum-weight weakly and strictly fundamental forward cycle basis problems. In this paper, we address these open questions. First, we study the existence, structure, and computational complexity of minimum-weight forward cycle bases. We give a complete structural characterization of digraphs that admit weakly fundamental (and hence integral) forward cycle bases. We further provide a characterization when a strongly connected digraph admits a forward fundamental cycle basis, proving that such a basis exists if and only if the set of directed cycles has cardinality equal to the cycle rank; in this case, the basis is unique. Lastly, we show that while minimum-weight forward fundamental cycle bases can be found in polynomial time whenever they exist, the minimum-weight forward weakly fundamental cycle basis problem is APX-hard via an L-reduction from the minimum-weight weakly fundamental cycle basis problem on digraphs with metric weights. Second, we introduce opt-in graphs, i.e., the family of graphs for which minimum cycle bases are integral for any weight function. We show that this family is minor-closed and hence, by the Robertson-Seymour theorem, is characterized by a finite set of forbidden minors, so that the opt-in recognition problem is solvable in polynomial time. Lastly, we present an algorithm to check whether a graph is opt-in, and if not, to identify which of its minors belong to the set of forbidden minors. Applying this algorithm, we show that the complete graph $K_n$ is opt-in if and only if $n \leq 7$.
Read the full abstract

Information Retrieval

Counter-GEO-Bench: Evaluating Defenses Against Information-Distorting Generative Engine Optimization

Generative engine optimization (GEO) enables content producers to increase the visibility of their web pages in generative search engines, but the same techniques can deliver targeted misinformation when adversaries publish ordinary-looking GEO-optimized documents that victim large language models (LLMs) retrieve and synthesize into distorted answers. No existing benchmark evaluates defenses against this threat under controlled conditions. Therefore, we present Counter-GEO-Bench, a defense benchmark that pairs 247 human-verified, quality-gated queries with information-preserving and information-distorting GEO rewrites, and evaluates defenses on attack success rate (ASR), false positive rate, and answer quality across three victim LLMs. Under Counter-GEO-Bench, three off-the-shelf defenses (Granite Guardian, Llama Guard 3, and NeMo Self-Check Fact-Checking) reduce ASR by at most 5.7% relative, while Granite Guardian's reduction is not statistically significant. Safety-taxonomy guardrails target policy violations, while GEO misinformation passes through them as fluent informational content. To this end, a lightweight benchmark baseline, C-GEO Guard, is proposed, reducing ASR by 47.6% relative with near-zero utility loss, which proves threat tractable.
Read the full abstract

Computation and Language

DiffIE: Diffusion-based Open Information Extraction

A single sentence often expresses multiple valid relational triplets, which makes Open Information Extraction (OpenIE) fundamentally a multi-output task. Existing neural systems handle this by autoregressive generation, which is flexible but slow and prone to redundancy, or by fixed-slot prediction, which is efficient but couples the extraction budget to training. We introduce DIFFIE which instead treats the stochasticity of conditional discrete diffusion as the extraction mechanism itself: independent reverse-diffusion trajectories over per-token role tags produce a pool of candidate triplets, which are clustered under lenient matching and ranked to form the output. Both the pool size and the number of returned extractions are inference-time choices, decoupling the extraction budget from training and exposing test-time compute as a tunable axis. DIFFIE achieves the new state of the art in CaRB (1-1) both F1 and AUC, and outperforms the strongest rule-based system (ClausIE) in BenchIE; it also remains competitive in standard CaRB and WiRe57 evaluations, giving the best average score among systems that report all four benchmarks. Ablations show that uniform discrete diffusion outperforms absorbing state diffusion in our setting, and that a matched non-diffusion stochastic tagger does not reproduce its gains. Our results indicate that diffusion stochasticity is an effective mechanism for structured prediction tasks with multiple valid outputs.
Read the full abstract

Solar and Stellar Astrophysics

Surface brightness inhomogeneity on the helium-rich hot subdwarf in the X-ray binary HD 49798

HD 49798 is an X-ray binary consisting of a hot helium-rich subdwarf primary and an accreting compact companion with an orbital period of 1.55 d. TESS observations of its optical light curve suggest short-period brightenings superimposed on a sinusoidal variation. The latter can be attributed to the tidal deformation of the primary, while frequency analysis of the additional variations shows up to eleven frequencies which form an harmonic series based on a period of 1.4 d, in excellent agreement with the stellar rotation period inferred independently from spectroscopy. These brightenings may be explained by a slowly changing and inhomogeneous temperature distribution across the surface of the hot subdwarf, such as might be associated with a magnetic field or with the presence of Rossby waves in the stellar envelope.
Read the full abstract

General Relativity and Quantum Cosmology

Redshift Spectroscopy as a Probe of Regular Black Holes, Black Bounces, and Scalar-Hair Compact Objects

Motivated by [Phys. Rev. D. 107, 064019 (2023)], we develop a unified and model-independent framework for the spectroscopy of photon frequency shifts in generic static, spherically symmetric spacetimes. Working with a line element characterized by three arbitrary radial metric functions, we derive exact expressions for the conserved quantities of massive and massless probes, the conditions for circular timelike geodesics, the local-emission-angle-dependent photon impact parameter, and the corresponding local redshift and blueshift branches measured by distant static observers. The formalism is further extended to include the line-of-sight peculiar motion of the source and the local propagation of photons in a nonmagnetized cold plasma, thereby identifying the gravitational, orbital, and dispersive factors entering the frequency-shift signal under the stated assumptions. We also show that, in vacuum, the same geometric structures governing orbital spectroscopy determine the photon sphere and the shadow impact parameter whenever an external null critical orbit is present. To make the framework suitable for deformed compact-object models, we construct a perturbative expansion around Schwarzschild geometry up to second order in a dimensionless deformation parameter, obtaining explicit corrections to the orbital energy, angular momentum, emitter four-velocity, photon impact parameter, and the vacuum and plasma frequency shifts. We apply the formalism to regular black holes from nonlinear electrodynamics, the Simpson–Visser black-bounce spacetime, and the Fisher–Janis–Newman–Winicour–Wyman geometry. These examples show that the same local spectroscopic language can be used across regular-black-hole, black-bounce, wormhole, and scalar-supported horizonless sectors.
Read the full abstract

High Energy Physics (Phenomenology)

Quark masses and mixing in the D_5 model under spontaneous CP violation

It is known that CP violation occurs in flavor physics, and the CKM matrix is complex. We investigate the Yukawa sector of a four-Higgs model based on D_5 symmetry. To understand all possible sources of CP violation in the Yukawa sector, we systematically analyze the impact of all possible representation assignments of the left-handed and right-handed quark fields under the D_5 group. After imposing the conditions of non-block-diagonal CKM matrix and the absence of massless quarks, we find that the only viable representation consists of one D_5 doublet and one singlet. In this case, two distinct types of quark mass models emerge in the quark sector. For each model, we introduce the most general vacuum that allows spontaneous CP violation (SCPV) into the mass matrices. Due to the specific structure of the D_5 symmetry, the vacuum expectation values (vevs) satisfy particular relations such that, for both quark models, the up- and down-quark mass squared matrices (i.e., M_d M_d^\dagger and M_u M_u^\dagger) can be simultaneously block-diagonalized by the same unitary transformation, forcing the CKM matrix to be block-diagonal and thus yielding vanishing mixing angles and CP violation. This result rules out the possibility of explaining experimental observations within the D_5 4HDM under the SCPV framework.
Read the full abstract

General Relativity and Quantum Cosmology

Boundary-Geometry-Driven Black Hole Formation in Vacuum

We identify a boundary-geometry-driven mechanism for the dynamical formation of marginally outer trapped surfaces (MOTSs) in vacuum general relativity. Mild three-dimensional anisotropies evolve inside a compact Cauchy domain whose effective isotropic thickness remains controlled, while the generalized boundary mean curvature can increase during either contracting or expanding boundary evolution. This drives the boundary across Yau's geometric threshold, forcing MOTS formation from initially untrapped data. We further interpret the characteristic shear construction of Ref.~\cite{MondalYau2026} as the null manifestation of the same anisotropic vacuum dynamics. The result provides a purely vacuum physical realization of MOTS formation through global geometric effects, without invoking a short-pulse concentration mechanism for gravitational radiation.
Read the full abstract

Dynamical Systems

Analytic rigidity and symbolic dynamics for two-centre billiards

We establish a sharp rigidity–chaos dichotomy for planar two-centre billiards, motivated by a natural analogue of the Birkhoff–Poritsky conjecture: the only tables integrable at every energy should be ellipses confocal with the two centres. Let $\Omega$ be a bounded domain with $\mathcal C^1$ boundary containing the segment joining the centres. At every fixed energy $h\geq 0$, if $\partial\Omega$ is not a confocal ellipse, we construct billiard trajectories that shadow the stable and unstable manifolds of the collision–reflection orbit and realise arbitrarily prescribed sequences of sufficiently large winding numbers around the segment. This yields an invariant set semiconjugate to the full shift on a countable alphabet, periodic trajectories with prescribed finite itineraries, and compact invariant subsystems with arbitrarily large topological entropy. If, in addition, $\partial\Omega$ is real-analytic, every real-analytic function on the fixed-energy phase space $M_h$ that is invariant under the billiard map is constant. This establishes the real-analytic form of the two-centre Birkhoff–Poritsky conjecture throughout the non-negative-energy regime.
Read the full abstract

Computation and Language

Efficient GUI Agents: A Systems Survey of Observation, Memory, Action, and Runtime Optimization

GUI agents increasingly operate across websites, mobile apps, and desktop environments, yet the field still reports progress primarily through task success. We argue that practical deployment depends equally on efficiency: how much context, computation, action budget, and runtime overhead an agent consumes while succeeding. This survey studies efficient GUI agents through an end-to-end systems lens that preserves the current technical axes of observation efficiency, context and memory efficiency, action efficiency, and planner-side/system efficiency. For each subsection, we expand the seed literature through targeted search plus backward and forward citation chaining, then synthesize the dominant mechanisms, reported efficiency signals, and new overheads they introduce. Across the literature, recent progress converges on a small set of recurring ideas: selective reading instead of full-context ingestion, global-to-local visual allocation, recoverable memory rather than raw history replay, verification-aware control, and hybrid runtimes that can switch between GUI and non-GUI execution. We conclude by identifying the main open problems, including honest accounting of verifier cost, cross-benchmark comparability, and co-design of observation, memory, and execution layers under real latency and privacy constraints.
Read the full abstract

All authors Bizhe Bai, Jiakang Yuan, Hongming Wu, Xinyue Wang, Jie Ren, Siyao Chen, Yuchen Ya, Fan Bai, Pai Peng, Huafeng Qin, Tao Chen

Number Theory

Affine Copies of Three-Point Patterns in Sets of Integers

Let $P=\{0,a,b\}$, where $0<a<b$ and $\gcd(a,b)=1$. For a finite set $A\subset\mathbb Z$, let $M_P^+(A)$ count the copies $x,x+ad,x+bd\in A$ with $d>0$, and let $M_P(A)$ count the copies with any $d\ne0$. We prove that every such three-point pattern other than the arithmetic progression $\{0,1,2\}$ satisfies \[ M_P^+(A)\le \frac{99}{400}|A|^2+O(|A|), \qquad M_P(A)\le \frac{13}{28}|A|^2+O_P(|A|). \] For the particular pattern $P=\{0,1,3\}$ – the subject of a question raised by Ganguly and recorded as Problem 24 in Green's list of open problems – we sharpen the bound allowing both signs of the dilation to \[ M_{\{0,1,3\}}(A)\le \frac{47}{122}|A|^2+O(|A|). \]
Read the full abstract

K-Theory and Homology

The decomposition relations alone present the Goncharov Lie coalgebra

Recently, Kupers–Rudenko–Sierra introduced the Goncharov Lie coalgebra. They gave a presentation using correlator generators and four types of relations: homogeneity, cyclic symmetry, shuffle relations, and decomposition relations. We show that the first three families of relations follow from the decomposition relations.
Read the full abstract

Robotics

Contact-Constrained Lower-Limb Joint-Offset Calibration for Humanoid Robots

Accurate joint encoder offsets are essential for kinematic consistency in humanoid lower limbs, yet existing calibration methods typically require external motion-capture systems or fiducial targets. We present a self-contained calibration framework exploiting only onboard joint encoders and a pelvis-mounted IMU during static double-support contact. The inter-foot transform from forward kinematics must stay constant when both feet are fixed; minimizing its posture-dependent dispersion yields a nonlinear least-squares problem over the 12-dimensional offset vector. A Hessian eigenstructure analysis shows that parallel pitch axes induce a rotational coupling. Orientation residuals then observe only the pitch-offset sum, while translation and posture diversity set the remaining numerical observability. For the A3 pitch-to-roll-to-yaw ordering, hip-roll and hip-yaw excitation reduce hip-pitch coupling. A standing-posture knee prior then anchors the remaining weak pitch-chain decomposition. Simulation and real-machine injection tests show consistent recovery, and on held-out recordings calibration reduces foot-height RMS residuals from 4.26 to 2.20 mm on A3 and from 8.03 to 1.43 mm on A2. An independent LiDAR-inertial reference checks the pitch-coupled channel. Removing an injected pitch offset moves the leg-odometry vertical drift back toward the LiDAR trajectory. A few static double-support stances thus provide contact-consistent corrections for well-excited directions. Individual offsets in the weak pitch chain remain prior-dependent.
Read the full abstract

All authors Kaixiang Lu, Haiyu Lan, Chunxiao Qiao, You Li, Chengyuan Luo, Enyu Li, Peiwen Lin, Chuang Wang

Signal Processing

Enhancing Vehicular Network Performance Through Integrated RSU and UAV Deployment

The increasing density of connected vehicles can place substantial pressure on fixed roadside infrastructure, particularly when the available communication resources become insufficient to accommodate temporary traffic surges. This paper investigates auxiliary unmanned aerial vehicle (UAV) assistance as a flexible mechanism for improving service availability and throughput in vehicular networks. Two representative network configurations are considered. In the first, an auxiliary UAV (UAVa) supplements two fixed roadside units (RSUs), whereas in the second, UAVa assists a heterogeneous infrastructure comprising one RSU and one UAV. Vehicle service is determined according to node coverage, the line-of-sight (LoS) probability of aerial links, and a prescribed signal-to-interference-plus-noise ratio (SINR) requirement. The resulting framework enables UAVa to accommodate eligible vehicles that cannot be adequately served by the primary infrastructure as the network load increases. Simulation results show that, under the considered configurations, the aerial nodes benefit from more favorable propagation conditions and achieve higher throughput than the fixed terrestrial RSU. Moreover, the introduction of UAVa increases the available service capacity under high vehicular loads, both for a purely terrestrial baseline and for a network already supported by an aerial node. These results demonstrate the potential of auxiliary UAV assistance as a flexible load-relief mechanism for capacity-constrained vehicular networks.
Read the full abstract

All authors Muhammad Ismail, Syed Luqman Shah, Fazal Muhammad, Zeeshan Shafiq

Machine Learning

Bayes-Optimal BER and AUC: Estimation and Evaluation of Estimators

A fundamental quantity in machine learning is the optimal performance achievable by any model on a given task. Estimating this quantity allows us to distinguish the irreducible part of the error from a deficiency of the model, telling us how much room for improvement remains. Recent work has shown that the Bayes error, or equivalently the optimal accuracy, can be estimated from soft labels in binary classification. However, accuracy is often a poor summary of performance in settings with severe class imbalance or noisy annotations, where metrics such as the balanced error rate (BER) and the area under the ROC curve (AUC) are more appropriate. We address this gap with two complementary contributions. (i) Estimation. We propose soft-label-based estimators for the optimal BER and AUC. We first consider the clean setting in which true soft labels and the class prior are known, and then extend the estimators to a more realistic setting in which the class prior is unknown and the observed soft labels are corrupted by an unknown order-preserving transformation, possibly followed by additive noise. In the latter setting, we approximately recover the clean soft labels via isotonic regression with auxiliary hard labels, estimate the class prior with a clipped mean of the hard labels, and derive finite-sample error bounds for the resulting plug-in estimators. (ii) Evaluation. Since the optimum is unobservable on real datasets, evaluating any such estimator is itself nontrivial. We extend the FeeBee framework, originally proposed for evaluating Bayes-error estimators, to the optimal BER and AUC. The resulting procedure provides practical evaluation scores without requiring knowledge of the optimum, and applies to any estimator of the optimal BER or AUC, not only our proposed ones. Experiments on synthetic and real-world datasets validate both the estimators and the evaluation procedure.
Read the full abstract

Information Retrieval

Genuine Information Needs of Social Scientists Looking for Data

Publishing research data is widely expected to increase its reuse and to inspire new research. In the social sciences, data from surveys, interviews, polls, and statistics are primary resources for research. There is a long tradition to collect and offer research data in data archives and online repositories. Researchers use these systems to identify data relevant to their research. However, especially in data search, users' complex information needs seem to collide with the capabilities of data search systems. The search capabilities, in turn, depend to a high degree upon the metadata schemes used to describe the data. In this research, we conducted an online survey with 72 social science researchers who expressed their individual information needs for research data like they would do when asking a colleague for help. We analyzed these information needs and attributed their different components to the categories: topic, metadata, and intention. We compared these categories and their content to existing metadata models of research data and the search and filter opportunities offered in existing data search systems. We found a mismatch between what users have as a requirement for their data and what is offered on metadata level and search system possibilities.
Read the full abstract

All authors Andrea Papenmeier, Thomas Krämer, Tanja Friedrich, Daniel Hienert, Dagmar Kern

Artificial Intelligence

Improving Evaluation Realism with Inference-Time Compute and Deployment Scaffolds

A core obstacle to alignment evaluation is evaluation awareness: capable models can tell when they are being tested rather than deployed, weakening the conclusions a safety evaluation can support. We present two techniques that make simulated alignment evaluations harder to distinguish from real deployments. Our first technique, critique refinement, spends additional inference-time compute on each simulator action: the simulator generates multiple candidate actions, refines them using feedback from an instance of the target model on how to make them more realistic, and continues the evaluation with the most deployment-like candidate. Our second technique, DISH (Deployment-Imitating SWE-Agent Harness), wraps the target in an agent harness, reducing the gap between simulated and real deployment environments in coding settings. We test the techniques on multiple target models and find that applying both yields larger realism gains than either alone. We show that automated approaches can improve the realism of alignment evaluations, and that they compare favorably to spending the extra compute on more audit attempts.
Read the full abstract

All authors Axel Ahlqvist, Richard Guan, Juan-Pablo Rivera, Adeline Kassler, Dmitrii Troitskii, Alexandra Souly, Kai Fronsdal, Robert Kirk, John Hughes

Fluid Dynamics

Improved uncertainty representation for reducing artificial energy production in structured input-output stability analysis

This work employs a new form of fixed structured uncertainty within the structured small-gain theorem approach proposed by Frank-Shapir & Gluzman (J. Fluid Mech., vol. 1030, 2026, pp A8) for the stability analysis of incompressible shear flows subject to finite-magnitude disturbances. Within this framework, the nonlinear advection term in the Navier-Stokes equations is replaced by a structured feedback uncertainty interconnection with the linearized dynamics to account for the impact of nonlinear feedback. Herein, a new uncertainty representation is derived via linear transformations of the input and output channels, transforming the feedback loop such that the resulting structured uncertainty has a repeated-diagonal structure. This structure aims to preserve the component-wise pathways of the nonlinear advection term while keeping the structured singular value computation tractable. We apply the method to two canonical base flows: Couette and plane Poiseuille flows. The resulting thresholds on disturbance magnitude to preserve stability are less conservative and more accurate. We compare the novel methodology presented here with previously proposed repeated and non-repeated block approximations of the uncertainty structure, where our stability threshold provided the closest agreement with previous numerical and experimental studies. We show that repeated and non-repeated block structures that were proposed in past studies result in an artificial energy-production term arising from using constant structured uncertainty in the structured input-output formulation, violating the divergencefree assumption. This energy-production term is smallest when using the methodology presented in this work, providing a more faithful representation of the impact of nonlinear feedback interconnection with the linearized dynamics of the Navier-Stokes system.
Read the full abstract

High Energy Physics (Theory)

Residues on Permanent Pinches: Finite Integrals and Leading Divergences

We study infrared (IR) divergences from the point of view of permanent pinches. We show how canceling permanent pinches allows us to form finite linear combinations of integrals. We further show how to compute scheme-independent leading IR divergences with- out introducing a regulator, by integrating certain residue forms along desingularizations of permanent pinch varieties. Our methods work for massless and mixed massive-massless inte- grals, for non-planar integrals with arbitrary numerators as well as for integrals with higher (integer) powers of propagators. We use this analysis to determine leading singular terms in some examples of box integrals. We further study examples of finite integrals at higher loops, and in particular provide a basis for finite non-evanescent integrals in a non-planar two-loop five-point topology.
Read the full abstract

Computer Vision

Domain shift-robust object detection with GenAI image editing

Object detectors often degrade under domain shifts such as changes in lighting, weather, or occlusion. These shifts alter object appearance and expose a reliance on visual shortcuts learned from the training distribution that do not generalize across domains. Acquiring sufficient real-world samples to capture such domain variation is particularly difficult in specialized, low-data settings. Recent advances in diffusion-based generative image editing have shown promise for improving the in-domain performance of object detectors through synthetic data augmentation. However, their potential to improve out-of-domain robustness remains largely unexplored. We hypothesize that generative image editing can simulate a controlled domain shift in training data, effectively bridging the gap between source and target domains. To test this, we studied camouflaged military vehicle detection as a challenging domain shift scenario. Detectors trained on uncamouflaged data demonstrate substantial degradation on real test imagery containing foliage, netting, and multi-spectral camouflage across 15 vehicle classes in close-up, ground-level imagery. We used two diffusion-based editing models, Qwen Image Edit 2509 and Flux.2 Dev, to synthetically add camouflage to the training data, alongside a LoRA fine-tuned version of Qwen. A non-generative black-bar occlusion baseline served as a lower bound on augmentation quality. Using a GroundingDINO detector trained on real and synthetic data, generative camouflage augmentation yielded substantial mAP improvements for foliage (+20.1) and netting (+14.4) camouflage. Generating multi-spectral camouflage proved more challenging, but LoRA fine-tuning improved performance by 4.4 mAP over the uncamouflaged baseline.
Read the full abstract

All authors Isabel D. Stein, Thijs A. Eker, Sebastiaan P. Snel, Ella P. Fokkinga, Klamer Schutte, Luca Ambrogioni, Friso G. Heslinga

Fluid Dynamics

Mechanism-Resolved Interface Momentum Transfer in Immersed-Boundary Lattice Boltzmann Simulations

Immersed-boundary lattice Boltzmann (IB-LBM) simulations of cylinder and particle flows are usually compared by ranking boundary-enforcement schemes, kernels, and collision models. Such comparisons can obscure the mechanism that controls each response. Here, canonical fixed-cylinder, oscillating-cylinder, and sedimenting-particle cases are used as controlled probes of Eulerian-Lagrangian interface momentum transfer. For prescribed bodies, differences among direct forcing (DF), multi-direct forcing (MDF), and distribution-function correction (DFC) are most clearly discriminated by local no-slip fidelity rather than by a universal drag ranking. For DFC, the kernel-dependent drag-coefficient ranking reversal between the hat and Peskin 4-point kernels is associated with the spatial redistribution of the marker-resolved correction and the resulting near-boundary slip and pressure deviation, rather than the total correction magnitude alone. For sedimenting particles, single-particle settling provides a moving-body baseline, whereas two-particle differential-density wake-interaction sedimentation shows the wake-exposed light particle to be comparatively more sensitive to the explicit internal-mass correction in the force evaluation, a configuration-dependent, finite-window moving-boundary closure response. Targeted two-relaxation-time (TRT) and central-moment multiple-relaxation-time (CM-MRT) collision controls remain secondary to the boundary and closure mechanisms within the tested regimes (prescribed-body comparisons at Re <= 200, two-dimensional, and the reported moving-particle cases). The resulting picture replaces a universal scheme ranking with a mechanism-resolved interpretation of IB-LBM interface momentum transfer.
Read the full abstract

Spectral Theory

Robin counting functions on cuboids in the critical scaling regime

We consider eigenvalue counting functions of Robin Laplace operators on cuboids where the Robin parameter and the spectral cut-off $\lambda$ are coupled. Our main focus is the critical regime, in which the Robin parameter is proportional to $\sqrt{\lambda}$. We obtain a two-term asymptotic expansion for the counting functions in this coupled setting. In the critical regime, the second term in the asymptotic expansion depends non-trivially on the proportionality constant and interpolates continuously between the corresponding second terms for the Dirichlet and Neumann Laplacians. Outside the critical regime, one recovers the corresponding Dirichlet or Neumann asymptotics. We also establish a Pólya-type inequality for the counting function whenever the ratio of the Robin parameter and $\sqrt{\lambda}$ exceeds a dimension-dependent threshold.
Read the full abstract

Computers and Society

Meeting the Coming Wave: The Emerging Politics of AI and Work across 33 Parliaments

A new politics of artificial intelligence and work is taking shape across party systems, but comparative politics has yet to map it. Using 1,514,950 parliamentary speeches from 33 parliaments (2023-2026), we show this politics follows a different logic than political economy expects. Research anticipates that technological disruption generates demands for compensation; instead, compensation accounts for just 2.3% of response-frame mentions, while enablement and investment dominate (55.2%), regulation and restriction follow (21.8%), and training (20.6%) appears at similar rates across families. Parties disagree instead over what AI means for work and how far this technology should be restrained. The mainstream and radical right support unrestricted enablement; the left is critical but divided on remedy. Social democrats stay adoption-oriented; greens split evenly. The radical left is the clearest force for restriction. The AI conflict thus concerns not compensation after disruption, but whether politics should enable technological change or govern its trajectory.
Read the full abstract

Superconductivity

Quantum Structural Renormalization and Anharmonic Stabilization of Superconductivity in P4/mmm YScH8

The recent synthesis of P4/mmm-YScH$_{8}$ at 191.7 GPa with a superconducting critical temperature Tc of 113 K has attracted considerable interest in the study of ternary superhydrides. Here we systematically evaluate the effects of quantum and anharmonic motion of ions in P4/mmm-YScH$_{8}$ using the stochastic self-consistent harmonic approximation (SSCHA). We find that these effects renormalize the crystal structure and lower the dynamical stability threshold pressure from $\sim$140 GPa to $\sim$84 GPa, a reduction of about 40%. Furthermore, neglecting these effects causes a significant overestimation of the $T_{c}$, for instance by $\sim$20 K at 190 GPa. Our prediction of $T_{c}$ at 190 GPa is 113 K ($\mu^{*}$= 0.13), close to the experimental observation. Analysis of the phonon density of states and Eliashberg spectral function further reveals site-dependent contributions to superconductivity. These results establish that a consistent treatment of nuclear quantum motion and anharmonicity is essential for accurately predicting superconductivity in compressed ternary hydrides.
Read the full abstract

All authors Xiaorui Dong, Mi Pang, Wenjie Ma, Yao Ma, Pugeng Hou

Analysis of PDEs

Existence and Blow-Up for Nonlinear Fokker-Planck with Controlled Drift Divergence

We extend the global existence results for critical parameters of [Bianchini and Leccese, J. Math. Anal. Appl. (2024)] to the multidimensional problem \begin{equation*} \partial_t u + \text{div } (b(t,x) u^{1+k}) = \Delta u \end{equation*} for $b:\mathbb R\times{\mathbb R}^d\to{\mathbb R}^d$ non-autonomous fields, in the case \begin{equation*} (\text{div } b)_-\in L^\infty_{\mathrm{loc}} \big([0,\infty);L^{p,\infty}({\mathbb R}^d)\big), \qquad p>d. \end{equation*} The proof of global existence follows the parabolic symmetrization approach of [Bandle, J. Analyse Math. (1976)] and is based on the study of the function \begin{equation*} m(t,s)=\int_0^s u^*(t,r)\,dr. \end{equation*}
Read the full abstract

Machine Learning

SEAL: Reinforcing Global Safety in Mixture-of-Experts through Shared Expert ALignment

Mixture-of-Experts (MoE) is a scaling architecture for large language models that activates only a small subset of expert modules per token, enabling massive parameter growth with nearly constant computation. Recent Hybrid MoE architecture adds shared experts to capture consistently useful representations, further improving stability and generalization. MoE now powers many flagship open-source and commercial models, yet remains vulnerable to adversarial attacks. Specifically, sparse routing introduces a structural vulnerability: MoE safety hinges on which experts are activated, and adversaries can subvert this selection through jailbreak prompts, malicious fine-tuning, and weight-level pruning of safety-critical neurons. Existing defenses primarily focus on hardening the router, but an adversary may still manipulate or bypass the routing trajectory due to the routing process's nondeterministic nature, thereby collapsing the defense. To cope with this problem, we first identify theoretically and empirically that shared expert, an always-activated component containing a small proportion of safety-critical neurons, can overcome the uncertainty of sparsely activated routing path and serve as a router-independent anchor to enhance global safety alignment. Based on this insight, we propose SEAL, a training-time parameter-efficient defense that produces a plug-and-play adapter attached to shared expert, and SEAL++, a variant that adds an orthogonal constraint preserving pre-existing safety subspaces during training. We evaluate SEAL and SEAL++ across six attack scenarios that combine three adversarial inputs (harmful prompting, jailbreak, malicious fine-tuning) with and without neuron pruning. SEAL reduces attack success rate (ASR) by up to 60%, at a capability cost of at most 1.4% on a five-benchmark average. Additionally, SEAL can seamlessly integrate with router-level ......
Read the full abstract

All authors Qingyu Meng, Yiwei Zha, Jiahuan Pei, Koen Hindriks, Herbert Bos, Min Chen

Artificial Intelligence

SCX Router: Streaming Zero-Shot Model Selection with a Decoder-KV Classifier and a Real-World Task Ontology

The rapid proliferation of large language models (LLMs) and the growing diversity of their applications presents a unique optimization opportunity: selecting the right model for the task, while optimizing for speed, cost, and quality at a per-task level. However, inference endpoints can vary widely in quality, price, latency, context support, tool use, domain expertise, and reasoning behavior. This heterogeneity makes manual heuristics difficult to maintain and unlikely to achieve consistently favorable speed–cost–quality trade-offs on their own. We introduce \router{}, a lightweight GLiClass-based router that assigns a suitability score to each inference-time model label without autoregressive generation. The released 0.6B-parameter checkpoint combines a Qwen3 decoder with a shallow bidirectional scorer. Its decoder-KV execution path preserves a text-only key–value cache across a session, encodes only new dialogue turns, and evaluates transient candidate-label tokens without adding them to the persistent cache. The same checkpoint also predicts task type, difficulty, reasoning mode, and expected output length, and supports custom zero-shot labels. For task generation, we construct a task ontology with 23 families, 115 task types, 345 routable subtypes, 1,173 synthetic examples, and an orthogonal axis of 30 domains. Using this structure, we generate 150,000 verifier-scored tasks and 15,000 open-ended tasks. We then train the Qwen3 decoder on these tasks, while explicitly separating learned request prediction from per-task policies for attributes such as eligibility, cost, cache reuse, safety, and sovereignty. Across six LiveBench subsets, the router outperforms the mean candidate; on the selected 1,000-task subset, it achieves an aggregate top-1 score of 0.707 versus 0.696 for the strongest fixed model, with benchmark-dependent gains.
Read the full abstract

All authors Ihor Stepanov, Aleksandr Smechov, Mykhailo Shtopko, Dmytro Vodianytskyi, Oleksandr Lukashov

Computer Vision

VoRTeC: Taming Foundation Flow for One-step Real time Video Compression

Ultra-low bitrate video compression still faces critical challenges: traditional neural video compression inevitably introduces blurring artifacts, while diffusion-based generative video compression suffers from excessive decoding latency and poor temporal consistency. To address these issues, we propose $\mathtt{VoRTeC}$, a Video Compression framework built upon a foundational flow model (Wan2.1). By compactly encoding latent video representations, predicting the positions of compressed representations along flow trajectories, and integrating multi-scale priors, $\mathtt{VoRTeC}$ enables the compressor to harness generative video flow priors effectively. Without accessing the parameters or gradients of flow matching networks, our framework achieves one-step decoding and reconstructions with high perceptual fidelity. Meanwhile, we maintain consistency across frame groups via tail-frame reuse and prior caching. Extensive experiments demonstrate that our method reduces bit consumption by 58% compared to prior diffusion-based approaches, with decoding speed boosted by 3 to 197 times: $\mathtt{VoRTeC}$ achieves a decoding speed of 13 FPS at 720p and 32 FPS at 480p.
Read the full abstract

All authors Yichong Xia, Qinhong Wu, Bin Chen, Jinpeng Wang, Zeyuan Chen, Haoqian Wang

Information Theory

Agentic UE-CoMIMO for 6G Terminals: From Virtual Antenna Augmentation to AI-Native Virtualization

End-user-centric collaborative MIMO (UE-CoMIMO) lets nearby devices form a virtual multi-antenna terminal to overcome the antenna limitations of individual user equipment. Extending such cooperation to communication, sensing, computing, and task-relevant information exchange requires a control layer that can interpret user intent, select cooperation mechanisms, and replan as conditions change. This article introduces Agentic UE-CoMIMO, in which device micro-agents, a smartphone or CPE hub agent, and edge/network agents coordinate device participation, relay modes, traffic splitting and duplication, compute placement, semantic-token exchange, and topology reconfiguration. Two system-level scenario studies on creator-centric live streaming and wearable-collaborative blind-spot sensing compare the proposed controller with capability-matched adaptive baselines. The results show that, by anticipating changes and preparing cooperation and fallback actions in advance, agentic control sustains high-quality streaming for longer and maintains blind-spot warnings through device outages. We also discuss the associated standardization, interoperability, trust, and validation challenges.
Read the full abstract

All authors Chao-Kai Wen, Yen-Cheng Chan, Lung-Sheng Tsai, Pei-Kai Liao, Geoffrey Ye Li

Computer Vision

If It Moves, Radar Knows: A Physics-Aware Radar Transformer for Class-Agnostic Moving-Object Detection

Detectors trained on closed-set annotations can miss rare moving objects outside the training taxonomy. Automotive radar provides category-independent Doppler motion cues and is less affected by adverse illumination and weather, but sparse, noisy returns hinder class-aware 3D box detection. Surface location and velocity remain useful for motion reasoning and collision avoidance when full box geometry is difficult to recover. We present the Physics-Aware Radar Transformer (PART), a fully sparse radar-only detector that predicts existence confidence, a representative surface point, and 2D ground-plane velocity for each moving-object hypothesis. Doppler-Aware Query Initialization (DAQI) replaces scene-independent learned queries with input-dependent proposals by clustering radar returns in position and velocity, easing query-object assignment in sparse scenes. Physics-Guided Cross-Attention (PGCA) incorporates radial-Doppler consistency and radar cross section (RCS) into query-point association. Uncertainty-aware supervision randomly masks ground-truth objects and assigns soft existence targets to ambiguous radar-supported queries, reducing reliance on exhaustive annotations. With only 1.1 million parameters, PART achieves a class-agnostic average precision (CA-AP) of 0.8827, a mean average surface translation error (mASTE) of 0.3188 m, and a mean average velocity error (mAVE) of 0.8084 m/s on nuScenes. It attains 0.9203 recall on rare and safety-relevant categories excluded from the standard evaluation and remains effective at night, in rain, and under severe occlusion. Inspection of apparent false positives shows that some predictions correspond to moving objects absent from the nuScenes annotations. Code and pretrained model weights will be publicly available at https://github.com/sunyinghao-uestc/PART.
Read the full abstract

All authors Yinghao Sun, Shuguang Li, Jinliang Shao, Tieshan Li