Peer Revenue Past citations are no guarantee of future returns

Papers

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

All of arXiv

Sunday 13 September

Dynamical Systems

Topological Expansivity and Shadowing for Anosov Diffeomorphisms

We show that every Anosov diffeomorphism is topologically expansive in the sense of [9]. We also prove that Lipschitz shadowing forces the stable and unstable bundles to be uniformly transverse, with a quantitative bound in terms of the shadowing constant. Finally, we give two examples on complete Riemannian manifolds: one is expansive but does not have the shadowing property, while the other has finite volume, bounded sectional curvature, Lipschitz shadowing, and orthogonal invariant bundles, but is not expansive with respect to its Riemannian distance.
Read the full abstract

All authors Z. Li, C.A. Morales, S. Romaña, Y. Yang

Dynamical Systems

Generalized Bratteli diagrams without probability tail invariant measures

The paper focuses on the following problem: given a generalized Bratteli diagram $B$, determine conditions under which the tail equivalence relation $\mathcal R$ does not admit a tail invariant probability measure on the path space $X_B$. We give several sufficient conditions, formulated in terms of the incidence matrices, that guarantee the nonexistence of such a measure. We also discuss a new spectral method for finding tail invariant measures for a class of stationary generalized Bratteli diagrams. This method allowed us to answer affirmatively an open problem formulated in arXiv:2212.13803.
Read the full abstract

Computer Vision

Sparsity-Adaptive Sharpness-Aware Minimization

Deploying deep neural networks in real-world settings requires models that are both compact and robust to common corruptions. However, at deployment-relevant high sparsity, standard pruning pipelines often degrade corruption robustness, and existing sharpness-aware training/pruning approaches provide limited robustness gains. We address this issue by introducing Sparsity-Adaptive Sharpness-Aware Minimization (SA-SAM), which derives a sparsity-dependent SAM/ASAM perturbation radius by keeping the mean absolute perturbation (an $\ell_1$-based proxy) approximately invariant as sparsity increases. As a simple complementary option, we evaluate Magnitude-Weighted Hessian (MWH), derived from a second-order removal-path analysis, yielding an importance proportional to $\mathrm{Diag}(F)_i\,|w_i|$, where $\mathrm{Diag}(F)$ is the diagonal empirical Fisher used as a curvature proxy in our implementation. Across CIFAR-10-C, CIFAR-100-C, and ImageNet-100-C, our approach achieved stronger corruption robustness than the considered pruning baselines at 80–90% sparsity, while preserving clean accuracy. We additionally quantify the robustness–throughput trade-off by reporting measured inference throughput under sparse execution at deployment-relevant sparsity levels.
Read the full abstract

Probability

Disorder transfer for the critical SK overlap law

We give a quantitative disorder comparison for the zero-field Sherrington–Kirkpatrick model at inverse temperature one. For any fixed number of replicas, replacing Gaussian couplings by independent couplings from a bounded symmetric variance-one law changes a bounded replica expectation by at most a constant times the Gaussian overlap second moment plus $N^{-1}$. A multiplicative comparison for nonnegative observables controls the tails. The proof uses Talagrand's positive-observable estimates and Yu-Ting Chen's fourth-order replica cancellation. Assuming the Gaussian critical overlap limit stated by Du and Huang, we obtain the same limit for the random quenched overlap measure in the iterated Wasserstein-2 topology. No separate moment assumption is needed for this transfer. Their additional exponential moment bound gives convergence of the rescaled spin-glass susceptibility in every finite Wasserstein distance. The coupling class includes fair signs, uniform disorder, and signs thinned at any fixed positive density.
Read the full abstract

Quantum Physics

Quantum ensembles as waves on state space

A quantum ensemble contains preparation weights that are lost in its average density operator. We show how optimal decisions read these weights as a wave on projective state space. Varying the input prior sweeps a surface of fixed Born overlap through the ensemble; the resulting response propagates according to the geometry of pure states and reconstructs their distribution in every finite dimension. Conditional preparations can fix this wave through contact, selecting the ensemble at a steering boundary. Positive propagation gives exact biased-qubit solutions and explicit parent measurements for noisy channels. When measurement records are combined, their state directions are averaged and the wave acquires a source whose growth measures the lost discrimination advantage. Preparation, inference and noise can thus be understood as operations on the same recoverable distribution, linking quantum correlations to measurement resolution.
Read the full abstract

Systems and Control

Probe Sets for Nonlinear Kalman Filtering: Multi-Center Gains and Covariance Recalibration

Kalman filters (KFs) choose the gain as the product of the state-measurement cross-covariance and the inverse of the innovation covariance. In nonlinear systems, the predictive distribution generally changes shape, and these covariances require approximation. Conventional nonlinear KFs approximate them around a single center, the predicted state. They can produce inaccurate gains and overconfident covariance estimates when one local approximation does not represent the measurement geometry across the uncertainty region. To address this issue, we introduce probe sets, small covariance-scaled collections of states at which the filter repeats its local approximation. We then select the gain that minimizes the average covariance reported by these local approximations. The same construction is used in our previously proposed covariance recalibration step to improve covariance consistency. The construction applies to different KF variants without requiring a specific approximation rule. For quadratic measurements, we derive conditions under which probing yields lower normalized true error variance than the corresponding single-center gain when prediction uncertainty is sufficiently understated. We apply the framework to four KF variants and compare the resulting filters with multiple baselines across 300 randomized setups for each of two systems. The results show that probing preserves typical accuracy while substantially reducing covariance inconsistency and the rare large-error trajectories that dominate the root mean square of per-run errors. The code is available at https://github.com/Shida-Jiang/Probe_KF.
Read the full abstract

Geometric Topology

On Nilpotent and Hypocentral Quandle Homomorphisms

We introduce a notion of nilpotency for surjective quandle homomorphisms, which relativizes the nilpotency of quandles recently defined by Darné. A surjective homomorphism is called nilpotent if its relative inner automorphism group is nilpotent relative to the inner automorphism group of the source. We show that this condition can equally be described through the relative transvection group and the relative adjoint group, and that it is equivalent to being a finite composite of covering homomorphisms. Extending the relative lower central series to transfinite ordinals, we then introduce hypocentral homomorphisms and characterize them as the surjections that are right orthogonal to the strongly connected homomorphisms. As a consequence, the strongly connected and the hypocentral homomorphisms form an orthogonal factorization system for the surjections in the category of quandles. Finally, we also introduce reduced and operator-reduced homomorphisms, which relativize reduced quandles, and give sufficient conditions for such a homomorphism to be nilpotent.
Read the full abstract

Analysis of PDEs

Direct and Indirect Physics-Informed Neural Networks for Dirichlet Boundary Control of Semilinear Parabolic Equations: A Conditional Error Analysis

We study physics-informed neural networks (PINNs) for the Dirichlet boundary control of a semilinear parabolic equation with Tikhonov regularization. Two approaches are considered. A direct PINN parameterizes the state and control by separate networks and minimizes a penalized form of the tracking objective. An indirect PINN instead represents the state, adjoint, and control by unconstrained networks trained jointly to satisfy the first-order optimality system, with the state-control coupling and the homogeneous adjoint boundary and terminal conditions imposed as soft penalty terms rather than enforced architecturally. For the indirect formulation we develop an error estimation framework that decomposes the total error into approximation, optimization, quadrature, and soft boundary/terminal-constraint contributions. Under standing assumptions on optimal-solution regularity and compatibility, network approximability, uniform Hölder control of the soft-constraint residuals, and a local neighborhood of the reference optimality-system solution, we derive a quantitative linearized stability estimate and a conditional local nonlinear residual-to-error estimate, and construct a computable residual indicator with a conditional reliability bound. Numerical experiments on two manufactured test problems - a cubic reactiondiffusion equation and a linear equation with a nontrivial boundary control and adjoint - illustrate the behavior of the direct and indirect formulations.
Read the full abstract

Robotics

Learning Communication-Conditioned Generative Policies for Decentralized Multi-Agent Collision Avoidance

In this work, we propose a decentralized communication-conditioned generative framework for multi-agent collision avoidance. Agents generate short-horizon action sequences using a flow-matching policy trained from privileged offline demonstrations with access to global state. The demonstrations do not include explicit communication signals; instead, agents learn to exchange and aggregate latent messages that encode interaction-relevant intent under partial observability. This formulation supports flexible inference at test time, where unconditioned generation corresponds to independent behavior and communication-conditioned generation enables coordinated interaction without centralized planning. The resulting policies operate in a fully decentralized manner at execution time, relying only on local observations and learned messages. Combined with a receding-horizon inference scheme, the proposed approach enables efficient single-step inference of short-horizon action sequences and degrades gracefully under communication dropouts. Extensive simulation results demonstrate near-expert collision avoidance performance and strong generalization to denser, unseen multi-agent scenarios, along with zero-shot transfer to real-robot experiments.
Read the full abstract

Computational Engineering

Public Opinion as an Option: Leveraging Prediction Markets to Hedge Exposure to Spot Crypto Volatility

This paper proposes an investment strategy through resource allocation into Kalshi Crypto Event Contracts in order to effectively hedge exposure to spot asset volatility. Using Bitcoin as a proof of concept, we treat corresponding Kalshi markets on the asset's future price as option contracts, and through construction of different portfolio allocations present a framework for which event contracts can be effectively utilized by retail investors in order to control risk framework without reducing underlying exposure to spot volatility. We showcase this information through analysis of three separate Kalshi event contracts, all relating to a different underlying market regime (i.e bullish, bearish, flat), and provide a comprehensive risk profile for each regime. Results show that a dynamic Kalshi/Bitcoin hedging strategy outperforms more simplistic portfolios across all regimes, supporting our initial hypothesis. We discuss further considerations which can erode profits in the future (i.e. fees), and provide direction for future research.
Read the full abstract

Probability

Sharp Norms from Finite Structure: Graph Matrices and Structured Chaoses

Graph matrices encode dependencies in random matrices built from shared random variables and arise in spectral algorithms, sum-of-squares (SoS), and high-dimensional statistics. We determine how finite graph structure controls their sharp spectral growth. For every fixed simple graph shape in the dense Rademacher model, including overlapping or empty matrix boundaries, we prove $\mathbb E\|M_\alpha\|=\Theta_\alpha(n^{(v+h-s)/2}(\log n)^{a_*/2})$, where $v$ counts vertices, $h$ isolated summation vertices, $s$ the minimum boundary-separator size, and $a_*$ maximizes an active-component count over minimum separators. Thus two finite cut optimizations determine both the polynomial and logarithmic exponents. The formula closes the polylogarithmic gap in separator bounds, and an infinite family with identical coarse parameters but different norms shows that the logarithmic exponent records genuinely new structure. The proof controls all defect layers in growing trace moments by converting label loss into separator excess; conditional flattening and synchronized fluctuations yield matching lower bounds. We extend the analysis to specified independent-factor chaoses, local weights, unequal dimensions, bounded asymmetric noise, Gaussian inputs, and fixed-degree Hermite inputs. Applications include degree-four clique SoS feasibility for $9\le k\le c\sqrt n$ without an asymptotic logarithmic loss, and Gaussian random tensor networks: deviation thresholds, sharp expected scales, entropy estimates, and, for connected loopless equal-dimensional networks, convergence of the rescaled largest output eigenvalue to the exact right edge of the limiting law. These results connect finite structure to sharp growth scales, and additional algebraic and spectral structure to full feasibility and exact limiting constants.
Read the full abstract

All authors Huibo Xu, Shi Fu, Youming Qiao, Dacheng Tao

Number Theory

Surjectivity of Finite Rank-Capped Enots-Wolley Sequences

For each fixed integer $K\ge2$, we consider the Enots–Wolley sequence in which every term after the initial $1,2$ is required to have between two and $K$ distinct prime divisors. We prove that every integer satisfying this restriction occurs. If an exact prime support $T$ were selected only finitely often, then after a finite cutoff the terms meeting $T$ would form short episodes, and every full term would force an earlier proper term at comparable numerical height. The case $|T|=K$ is then ruled out directly. In the remaining case $|T|<K$, at a record proper value $H$, greediness forces every unblocked rank-$K$ integer below $H$ containing exactly one prime of $T$ to have occurred earlier. Fixed-order Landau estimates show that this proper population has order $H(\log\log H)^{K-2}/\log H$, while the entire possible full population on the same scale has strictly smaller logarithmic order. This contradiction proves surjectivity. The theorem concerns each fixed rank cap and does not settle surjectivity of the unrestricted Enots–Wolley sequence.
Read the full abstract

High Energy Astrophysical Phenomena

Exploring the dynamics of the Coma galaxy cluster by mapping its X-ray emission line profiles with XRISM

The intracluster medium (ICM) in merging galaxy clusters exhibits turbulence and bulk flows. Unraveling these components is crucial not only for elucidating the geometry of the cluster mergers, but also for understanding the physics of magnetic field amplification and relativistic particle acceleration. XRISM/Resolve data for two $3'\times3'$ fields in the core of the Coma cluster reveales that the ICM in the central field moves with $\Delta cz = -430$~km~s$^{-1}$ relative to the cluster galaxy average, while that in the southern field moves with $\Delta cz = -730$~km~s$^{-1}$ (see \cite{2025ApJ...985L..20X}, hereinafter “Paper I”). In this paper, we perform a more detailed analysis of these data sets to search for non-Gaussian features in the Fe-K line complex profiles. In the spectra from the northwest (NW) quadrant of the central field, in addition to the main and redshifted ICM components ($\Delta cz = -40$ km s$^{-1}$) reported in Paper I, we find evidence of another, blueshifted component, moving with $\Delta cz = -1250$ km s$^{-1}$. For a systematic search for other significant velocity components, we perform a bias-free 3 eV step multi-component fit to the Resolve full-array spectra from the central and southern fields. This search uncovers another redshifted component in the southern field, moving with $\Delta cz \sim +1230$ km s$^{-1}$. We estimate the energy densities of the ICM turbulence and bulk motion to be similar to each other and several times greater than the energy density of the cluster's $B\sim 5~\mu$G magnetic field.
Read the full abstract

All authors Kosei Sakai, Kazuhiro Nakazawa, Maxim Markevitch, Dominique Eckert, Yuichiro Ezoe, Yuto Ichinohe, Richard Kelley, Richard Mushotzky, Nobuhiro Okabe, Yuki Omiya, Naomi Ota, Cicely Potter, Andrew Szymkowiak, Shunsuke Torii, Nhut Truong, Ayşegül Tümer, Yuusuke Uchida, Dan Wik, Irina Zhuravleva, John ZuHone

Theoretical Economics

Very Credible Auction

We introduce a notion and motivation for very credible auctions. This is a strengthening in the commitment in the credible auctions setting, making it more suitable for a modern digital applications.

Machine Learning

Bayesian optimization with kernel ensembles and disagreement-based acquisition for source localization and acoustic inversion

Joint source localization and geoacoustic inversion requires optimizing an objective built from an expensive normal mode propagation model. Bayesian optimization (BO) with a Gaussian process (GP) surrogate can obtain accurate parameter estimates within a limited number of forward model evaluations, but its performance depends on the choice of kernel family. With few observations in a seven-dimensional search space, no single kernel can be expected to perform consistently well across individual inversions. To reduce this dependence, we use a weighted ensemble of GPs with different kernel families, allowing the surrogate to adapt to the observed objective without committing to one kernel in advance. The ensemble is combined with an optimum-conditioned acquisition function that determines where the expensive objective should be evaluated next. Experiments on simulated and measured SWellEx-96 data show that the resulting method achieves the lowest mean final objective among the considered BO strategies and reduces parameter estimation error on most coordinates. Ablation results further show that the ensemble provides robustness to kernel choice, while the acquisition function accounts for most of the optimization gain.
Read the full abstract

All authors Heng Zhang, Haotian Xiang, Florian Meyer, Qin Lu

Robotics

VGFM: Expressive Robot Policies via Dense Value Guidance in Flow Matching

Recent robot learning paradigms increasingly rely on large offline datasets of robotic interactions to train control policies. Expressive generative models enable rich and multimodal action representations, expanding the capability of this paradigm for complex robotic control. However, policy improvement with multi-step generative actors remains challenging. In offline reinforcement learning (RL), incorporating value-based objectives along generative trajectories often introduces substantial training complexity, including backpropagation through time (BPTT), auxiliary architectures, or distillation losses. We propose Value-Guided Flow Matching (VGFM), a scalable offline RL framework that enables dense value-guided shaping within a flow-based policy while avoiding BPTT and additional algorithmic overhead. VGFM parameterizes the policy as a conditional flow-matching model in action (x-prediction) space, ensuring that each intermediate flow step produces a valid robot action that can be directly evaluated by a standard offline RL critic. This design allows value guidance to be applied at randomly sampled flow times without differentiating through the entire generative trajectory, while preserving inference-time flexibility by varying the discretization of the underlying flow ODE without retraining. Evaluated on robotic locomotion and manipulation tasks in OGBench, VGFM achieves strong performance across a wide range of tasks under rigorous evaluation protocols. With minimal hyperparameter tuning, these results demonstrate that VGFM provides a simple, scalable, and effective approach for expressive policy learning in long-horizon, goal-oriented robotic control.
Read the full abstract

Systems and Control

Recursive Self-Improvement LLM Agents for Inverter Dynamic Model Identification

This is a position paper. We demonstrate that recursive self-improvement (RSI) large language model (LLM) agents are a natural search engine for dynamic model identification of inverter-based resources (IBRs) whose internal controls are often proprietary and hidden from grid operators. White-box models provide physical transparency but require vendor disclosure; black-box models avoid this requirement but sacrifice interpretability; and existing grey-box approaches, including sparse and symbolic regression, are poorly suited to discovering feedback control architectures or incorporating control-engineering priors. Our position is that this gap can be alleviated by (1) restricting the search space to a typed vocabulary of standard control modules, including PI controllers, phase-locked loops (PLLs), low-pass filters, etc., composed under block-diagram grammar rules, and (2) using an RSI LLM agent to perform program search over candidate block-diagram models, guided by measured frequency-domain admittance data at the point of common coupling (PCC), while fitting the free parameters of each candidate by nonlinear least squares. We instantiate this position by adapting ThetaEvolve, an open-source program-evolution framework supporting in-context evolution and test-time learning, to inverter model discovery. In a proof-of-concept study on a grid-following (GFL) inverter benchmark, the RSI loop reduces the normalized root mean square error (NRMSE) of a naive open-loop model from 0.470 to 0.0435 and identifies a 15-module closed-loop structure that closely resembles the hidden ground-truth GFL controller.
Read the full abstract

All authors Jie Feng, Xiaoyang Wang, Xin Chen, Yuanyuan Shi

Atmospheric and Oceanic Physics

PepC-Global: A Basin-Tuned Probabilistic Tropical Cyclone Model with Enhanced Out-of-Sample Skill and Climate-Sensitive Over-Land Decay

We present PepC-Global, a global version of the Princeton environment-dependent probabilistic tropical cyclone (PepC) framework that uniquely implements basin-wise tuning within a single unified tropical cyclone climatology model. PepC-Global represents tropical cyclone climatology as a coupled stochastic process linking genesis, track, and intensity conditioned on large-scale environmental predictors. Each of the genesis, track, and intensity modules outperforms widely used linear models in out-of-sample tests. The intensity module also incorporates an environment-dependent over-land decay model, offering greater sensitivity to climate change signals than conventional fixed decay rate approaches. Systematic evaluation against observations demonstrates that PepC-Global closely reproduces genesis basin-wise frequency, seasonal cycles, interannual variability, and spatial distributions. The model also accurately captures basin-wise track patterns, along-coastline landfall frequency, and intensity statistics including lifetime maximum intensity and landfall intensity distributions. PepC-Global provides a versatile tool for probabilistic tropical cyclone hazard and risk assessment and a practical framework for investigating changes in tropical cyclone activity across future climate scenarios.
Read the full abstract

Computer Vision

SPARK: Representation-Level KV Memory Alignment for Safer Vision-Language Models

Vision-language models (VLMs) remain vulnerable to jailbreaks that distribute harmful intent across text and images, making unimodal safety mechanisms insufficient. We investigate whether this vulnerability can be mitigated directly in the multimodal key-value (KV) memory formed during prefill, without modifying model parameters at inference time. We introduce SPARK, a two-stage framework for targeted KV-memory repair. Stage 1 uses a disposable diagnostic adapter to identify harm-associated directions in multimodal key and value representations. Stage 2 projects out these directions, learns a lightweight residual repair, and anchors repaired keys with an image-structural prior to preserve visual grounding. Rather than applying the intervention uniformly, SPARK mixes repaired and original memory using a head-wise coefficient g_h* determined by intervention-relevant subspace energy E_h, requiring no explicit harm classifier at inference. Across LLaVA-OneVision-7B, Chameleon-7B, Qwen2-VL-7B, and InternVL2-4B, SPARK reduces multimodal attack success while preserving general capability. On LLaVA-OneVision-7B, image-only jailbreak attack success falls to 4.7%, while MMMU remains within 0.6 points of the undefended model (47.8 vs. 48.4) with near-baseline language quality. On MM-SafetyBench, attack success decreases from 39.2% to 12.4%. Even under white-box adaptive joint prompt-image attacks, attack success is limited to 20.3%, compared with 54.6% for the undefended model. These results suggest that multimodal jailbreak behavior can be substantially mitigated by selectively repairing intervention-relevant KV subspaces at prefill, particularly when harmful evidence is carried by the visual modality.
Read the full abstract

Computation and Language

DenMark: Robust Semantic Watermarking for Diffusion Language Models

Semantic text watermarks encode signals in meaning rather than surface token choices, offering robustness to paraphrasing and other semantic-preserving edits. Existing semantic watermarking methods are primarily designed for autoregressive language models (ARLMs), where completed candidate units can be generated and scored before generation proceeds. This paradigm does not naturally extend to diffusion language models (DLMs), where semantic units remain incomplete during intermediate denoising steps and tokens may be updated in flexible orders. We propose DenMark, a semantic watermarking framework that injects key-dependent signals directly into the DLM denoising process. DenMark partitions the output into fixed token regions and uses temporary rollouts as semantic lookahead: conditional completions estimate the eventual semantics of an incomplete region, enabling DenMark to select local updates with higher estimated semantic watermark scores. Repeating this procedure across denoising steps progressively accumulates watermark evidence in the final output. For detection, DenMark uses calibrated scanning over candidate unit sizes to remain robust to boundary shifts introduced by semantic attacks. Across four DLMs, three datasets, and four semantic attacks, DenMark gives the strongest overall robustness and achieves the best TPR at each reported operating point in 46 of the 48 backbone-dataset-attack combinations. Overall, DenMark provides an effective semantic watermark for DLMs. Code is available at https://github.com/TianhaoMa5/DenoMark.
Read the full abstract

All authors Tianhao Ma, Weihao Xuan, Dong-Dong Wu, Farshid Nooshi, Takashi Ishida, Gang Niu, Naoto Yokoya, Masashi Sugiyama

Computation and Language

Document Topic Alignment Metrics for Evaluating Topic Models of Short-Text Public Health Communications on Social Media

Topic models are widely used to analyze public health-related social media short texts, yet their evaluation remains dominated by metrics that focus entirely on generated topics alone. There is a lack of metrics that quantitatively assess whether assigned topics meaningfully represent the corresponding short-text posts. We propose Document-Topic Alignment metrics (DoTA), an assignment-aware evaluation framework comprising metrics that measure semantic alignment between documents (posts) and their assigned topics. We also introduce margin-based and discriminative variants that capture topic assignment confidence and distinguishability. We evaluate DoTA across five topic models on three public health-related social media datasets from X and compare DoTA metrics with conventional topic-based metrics. Results show that DoTA provides complementary evaluation cues and aligns meaningfully with human evaluations. These findings establish the need for assignment-aware evaluation and demonstrate that the addition of DoTA enables a more comprehensive and practically meaningful evaluation for assessing short-text topic modeling performance.
Read the full abstract

All authors Wangjiaxuan Xin, Shuhua Yin, Yaorong Ge, Shi Chen

Databases

Towards Anticipatory Databases Through Shared Data and Workload Semantics

Database management systems increasingly serve dynamic and exploratory workloads, yet many of their decisions still rely on low-level signals such as recency, frequency, and address locality. These signals capture how data was accessed, but not what is being examined or how an analytical focus evolves. We argue for treating workload semantics as a first-class control signal for anticipatory decision making. Central to this view, we introduce semantic locality and semantic trajectories, which capture relationships among nearby queries and how those relationships evolve across a session. We propose a framework that represents semantic context at the data, query, and session levels, models its evolution over time, and translates it into task-specific utility estimates. We instantiate this framework in semantic prefetching and semantic cache eviction, which share a semantic layer to make two separate decisions. Prefetching uses semantic trajectories to anticipate future accesses beyond what address-based locality can capture, while eviction uses semantic relevance to inform block replacement. These systems provide initial evidence that shared semantic context can support multiple DBMS components. We further outline how this principle can extend to other decisions and data systems, and discuss key challenges in representation, cost, adaptation, and evaluation.
Read the full abstract

All authors Farzaneh Zirak, Kasper Overgaard Mortensen, Farhana Choudhury, Renata Borovica-Gajic

Fluid Dynamics

Embracing Flow Unsteadiness: A High-Throughput Learning Platform Enables Vortex-Exploiting Bioinspired Propulsion

Biological swimmers and flyers exploit unsteady vortices for propulsion, whereas engineered vehicles usually suppress them as disturbances. Learning such flow exploitation in machines is difficult because real-fluid interaction data are scarce and unstructured exploration is unstable in high-dimensional, history-dependent flows. Here we present REEF, a co-designed physical-learning framework that integrates SHOAL, an eight-channel high-throughput array for real fluid–structure interaction, with V-STAR, a staged algorithm that converts these interactions into policies through imitation, offline internalization, and online adaptation. Across lift-based, drag-based, and momentum-jet propulsors, REEF expands the attainable force envelope to more than twice that of parameterized search. Particle image velocimetry shows that these gains arise from coordinated vortex formation, growth, and force projection, rather than refinement of a fixed motion-to-force mapping. Force-trained policies transfer zero-shot to free-moving robots whose body motion changes the surrounding flow, suggesting that REEF learns transferable wake-coupling principles for embodied propulsion in unsteady fluids.
Read the full abstract

All authors Fei Han, Xinyu Cui, Zhipeng Wang, Ning Yang, Hang Xu, Haifeng Zhang, Zhongming Hu, Jun Wang, Junfeng Du, Dixia Fan

Networking

Breaking the Duplex Barrier: Lane-Granularity OCS Scheduling for LLM Training

Optical circuit switch (OCS) can reconfigure physical connectivity to match the predictable communication schedules of large language model (LLM) training. Although each OCS light path is physically simplex, existing demand-aware OCS schedulers allocate capacity in duplex-port pairs, forcing equal bandwidth in both directions and stranding capacity under asymmetric node-pair traffic. This paper present LACE, the first offline OCS schedule compiler that independently allocates transmit (TX) and receive (RX) lanes for LLM training. Without changing the selected collective algorithms, operation order, or rank placement, LACE reconstructs directed node-level demand, jointly determines which consecutive operations share a configuration and how many simplex circuits serve each direction, and realizes these allocations as physical lane bindings and optical paths under per-node lane-inventory and multi-OCS fabric constraints. Software acknowledgments carry feedback over independently provisioned return paths, while coordinated link configuration and recovery verify each configuration before communication resumes. On a separate three-server testbed using fixed topologies and matched per-port rate limits, LACE's asymmetric connectivity achieves $1.80\times$ speedup for communication replay and $1.27\times$ for GPT-2 training over a symmetric-topology baseline. At larger scale, simulations of LLaMA-3.1 70B and 405B schedules with sixteen 400-Gb/s ports per server show that LACE achieves $1.21$–$2.04\times$ communication speedup over the latest duplex OCS scheduler.
Read the full abstract

All authors Bangbo Liang, Yupeng Chen, Sicheng Zhao, Peihao Huang, Di Yang, Bohua Xu, Bin Yang, Shizhen Zhao, Guo Chen

Quantum Physics

Efficient expectation value estimation for quantum circuits via extended stabilizer frameworks and adaptive variance estimation

Classical simulability of quantum circuits plays a central role in characterizing and quantifying quantum computational advantage. Typically, the presence of non-Clifford gates introduces an exponential runtime overhead for classically estimating expectation values of observables. In this work, we introduce an efficient simulation framework that integrates the sum-over-Clifford method with quasi-probability-based stabilizer simulation to directly estimate expectation values. To optimize efficiency, we incorporate a circuit-adaptive reduction via variance estimation protocol, which terminates simulations once an empirical confidence bound is sufficient to guarantee the prescribed error tolerance. Our method circumvents overly conservative Hoeffding's inequality, thereby reducing the dependence on the gate-wise stabilizer extent from quadratic to linear in the low-variance regime, while maintaining the same space complexity as previous approaches. We explicitly prove the sample-complexity reduction for random quantum circuits and T-doped Clifford circuits and also empirically validate these advantages in the quantum approximate optimization algorithm and quantum kernel method.
Read the full abstract

All authors Yunseo Hwang, Giwon Song, Kyoung Keun Park, Hyukjoon Kwon, Taehyun Kim

Neurons and Cognition

Nonlinear dynamics of random neural networks with second-order synaptic motifs

Classical theories of random neural networks typically assume independent connectivity, overlooking the local motif structures prevalent in biological circuits. Here, we investigate how four second-order synaptic motifs (chain, reciprocal, convergent, and divergent) shape the dynamics of nonlinear firing-rate networks. While previous studies have established that chain correlations generate outlier eigenvalues, we demonstrate that these motifs also jointly reshape the Jacobian eigenvalue bulk. Using the path-integral formalism, we derive a dynamic mean-field theory which reveals that the chain motif acts as a retarded feedback of the ensemble-mean activity through the response kernel, producing a rich repertoire of dynamical regimes, including ferromagnetic states and limit cycles. At sufficiently large magnitude, negative chain correlations produce a glassy, multistable regime that was previously mainly associated with partially symmetric networks. Our theory also distinguishes convergent from divergent motifs: divergent correlations primarily rescale temporal noise, while convergent correlations suppress temporal chaos by converting nonzero mean activity into quenched heterogeneity. Finally, analyses of the Lyapunov spectrum and participation-ratio dimension show that motif structure changes the geometry of chaotic activity, reducing entropy production and attractor dimensionality even when the effective spectral edge is held fixed. Together, these findings establish second-order motifs as a fundamental structural mechanism governing the dynamical regimes of local cortical circuits.
Read the full abstract

Analysis of PDEs

Stability of non-isentropic gaseous stars

We study the linear stability of compactly supported static spherical equilibria of the non-isentropic Euler–Poisson system. In the Schwarzschild-stable case, the linearized equations are realized as a separable Hamiltonian system on weighted spaces adapted to the physical vacuum. We prove stability against non-radial perturbations and show that the algebraic dimension of the radial unstable subspace is the Morse index of a density quadratic form subject only to the mass constraint. Thus the infinitely many linearized entropy constraints reduce, for radial perturbations, to one mass constraint and an explicit entropy reconstruction. We apply this criterion to two entropy prescriptions. No smallness condition is imposed on a fixed entropy–density relation: under the stated effective-pressure and branch hypotheses, a simple mass maximum is not a stability transition. By contrast, for a small entropy distribution fixed as a function of enclosed mass, the turning-point principle holds. In the Schwarzschild-unstable case, we construct the self-adjoint velocity operator from its quadratic form, prove a strictly negative spectral bottom and sharp exponential growth, and reconstruct energy solutions of the original first-order linearized system.
Read the full abstract

Optimization and Control

A hybrid crossover kangaroo escape optimization framework for engineering optimization and UAV path planning

Complex engineering optimization problems are often characterized by multimodality, high dimensionality, and nonlinear constraints, posing significant challenges for efficient and reliable computation. To address these challenges, this paper develops a hybrid crossover-based optimization framework that enhances population interaction and improves search efficiency. The proposed framework integrates two complementary mechanisms, namely a Levy long jump crossover strategy for global exploration and a horizontal-vertical crossover strategy for effective information exchange and local refinement, thereby improving convergence behavior and robustness. To support scalable computation, the method is implemented within a unified multi-backend computational framework based on FEALPy, enabling consistent and efficient execution across heterogeneous platforms, including NumPy and PyTorch on both CPU and GPU. This design enhances portability, reproducibility, and computational efficiency in large-scale optimization tasks. Extensive experiments on the IEEE CEC2022 benchmark suite demonstrate that the proposed framework achieves competitive or superior performance compared with several representative metaheuristic algorithms, as validated by Wilcoxon rank-sum and Friedman statistical tests. In addition, the method shows strong performance on constrained engineering design problems. Finally, the proposed framework is applied to UAV path planning, formulated as a constrained optimization problem, demonstrating its effectiveness and scalability in complex engineering scenarios.
Read the full abstract

Digital Libraries

ATTRICITE: Training an Open 4B Model for Citation Recovery toward Faithful Attribution

Faithful citation attribution begins with identifying the intended source for a scientific claim. We study this source-identification capability through citation recovery: recovering the paper cited by the original author from a citation-bearing passage. Our evaluation adopts the published author's citation as an observable human attribution signal and uses target recovery as a proxy for progress toward faithful attribution. We introduce ATTRICITE, an open 4B-parameter model trained for tool-using citation recovery within the CiteGuard retrieval environment, together with CITEALIGN, a 7,607-instance computer-science dataset drawn from recent scientific literature. For controlled evaluation, we construct a 709-instance benchmark subset of CITEALIGN, comprising 410 development instances from 2024 publications and 299 temporally held-out test instances from 2025 publications. Across three runs at an inference temperature of 0.7, GRPO fine-tuning improves Qwen3-4B from 49.4%$\pm$1.5% to 59.8%$\pm$0.2% target-match accuracy, a gain of 10.4 percentage points. Despite using only 4B parameters, ATTRICITE outperforms gpt-oss-20b and comes within 3.9 points of GPT-5.4-mini, while Gemma 4 31B IT achieves the strongest overall performance at 72.0%$\pm$1.0%. We release the model and collection pipeline https://github.com/KathCYM/AttriCite to support reproducible research on citation recovery toward faithful attribution in a continually evolving scientific literature.
Read the full abstract

All authors Yee Man Choi, Xuehang Guo, Songcheng Cai, Yimu Wang, Yi R. Fung, Qingyun Wang

Computer Vision

What Input Resolution Is Required for Bird Species Identification, and What Is Its Latency Cost on an Edge Device? A Study of 14 Input Resolutions and Six Architectures with On-Device Measurements

Bird-strike mitigation at wind farms requires identifying distant birds that span only tens of pixels, so the classifier's input resolution N is a design variable, not a fixed specification. We study it with a factorial design over 14 side lengths N (16 to 224), six architectures, two training and evaluation regimes and 30 random seeds – 2,520 checkpoints and 5,040 evaluations – plus latency measured on an NVIDIA Jetson Orin Nano. Four results. (1) The selected N depends on the target: 0.90 is met on validation by ResNet50 at N=112 in an estimated 1.85 ms (0.8980 on test) and 0.95 by DINOv2-L at N=144 in 12.70 ms; changing the model buys more accuracy than raising N (+5.93 versus +2.33 points at N=112). (2) The benefit of lowering N depends on the assumed preprocessing path: N=224 -> 80 saves 13.5% when each individual is decoded from its own file but 46.2% when the detector decodes the 4K frame once; the Pareto set grows from 20 to 23 configurations. (3) Accuracy must be measured on the deployed engine: half precision costs ViT-S/16 alone 4 to 7 points at N>=96 while the CNNs stay within 0.1 points, and with selection held at validation the choice differs at 26 of 176 targets. A broken FP16 engine can run faster than a correct one, undetectable from latency; admitting 14 ViT-S/16 FP32 configurations moves the recommendation over the 0.931-0.938 band and under the 10 ms budget. (4) ViT-L-scale models fit this device, but activations exceed the FP16 range; splitting the graph at transformer-block boundaries confines FP32 to the affected segments, making the deployed DINOv2-L chain 1.85x faster than the single-engine build. We also quantify how the regime-difference sign stabilises with seed count; a sensitivity split removing some forms of group sharing preserves all 14 non-trivial signs at the selection boundary.
Read the full abstract