Fluid Dynamics
Papers
Fresh from arXiv, typeset properly. Highlighted words trade on the market.
All of arXiv
Tuesday 1 September
We study the inverse energy cascade of two-dimensional turbulence by means of high-resolution numerical simulations forced at intermediate wavenumbers. To clarify the role of vortices generated near the forcing scale, we decompose the vorticity and velocity fields into vortex and background components. Although the vortex component carries only a small fraction of the total energy, it has a clear effect on the energy spectrum and produces deviations from dimensional Kolmogorov scaling, which is realized by the background flow. We derive a generalized von Karman–Howarth balance for the decomposed fields and obtain a relation involving the third-order correlation between vortex and background velocity increments. The simulations support this prediction, which we interpret as a constant inertial-range flux associated with vortex-background correlations.
Artificial Intelligence
Can LLMs Discover Scientific Laws in Real and Parallel Worlds?
Scientific law discovery has long been central to scientific progress, proceeding through iterative cycles of generating hypotheses, testing them against empirical evidence, and refining them under scientific constraints. As large language models (LLMs) become increasingly involved in scientific research, whether they can discover scientific laws and how to evaluate this ability remain open questions. A central evaluation challenge is to move beyond familiar published equations while keeping discovery tasks grounded in scientific data and constraints. We introduce SciLaws-Bench, a curated collection of scientific task packages grounded in the source literature, each linking a scientific problem, supporting data, published reference equations, and scientific-validity rubrics. Through agent-assisted curation and human verification, we assemble 118 problems spanning six disciplines, drawing on 381 papers, 291 candidate laws, and roughly 8M data points. Each problem supports two complementary evaluation settings. SciLaws-Real uses fixed scientific data to evaluate proposed laws for held-out predictive fit and scientific validity. SciLaws-Parallel evaluates recovery of a newly synthesized structural variant of a published equation through active queries to a simulator calibrated to the source data. Our evaluation reveals three limitations: good predictive fit need not imply scientific validity, recovering a published formula does not establish recovery of its new structural terms, and candidate selection remains a bottleneck in scientific law discovery. Project page: https://yiyihum.github.io/SciLaws-Bench
Read the full abstract
Computer Vision
What, Where, and How: Probing Spatiotemporal Representations in Video Foundation Models
Self-supervised video foundation models learn rich spatiotemporal representations, yet it remains unclear what visual concepts these representations encode, where they emerge across transformer layers, and how they are geometrically organized. In this work, we tackle these three questions through a systematic layer-wise analysis of V-JEPA 2 and VideoMAE-v2. We leverage lightweight probes trained to discover three temporally grounded properties: (i) camera motion understanding, (ii) intuitive physics, and (iii) anomaly detection. Both models encode camera motion, with best results ($>90$ ROC AUC) emerging at 60-70% of network depth, and achieve moderate anomaly detection performance ($>60$ ROC AUC), but remain near chance on intuitive-physics tasks, suggesting a limited encoding of deeper physical reasoning. Beyond classification, we find that temporal features from individual videos form smooth low-dimensional trajectories in representation space, suggesting that camera motion is not only linearly decodable but also geometrically organized. Based on these results, we apply geometry-aware spline-based steering in the model's latent representations to interpolate camera motion, yielding steered videos with smoother trajectories and more coherent temporal progression than linear interpolation.
Read the full abstract
Machine Learning
Linear Reusable Neural Bases Architecture for Network Compression
Memory constraints remain a critical bottleneck in the deployment of large-scale AI models. Parameter sharing across network depth reduces model storage, but repeatedly applying an identical transformation limits flexibility across layers. Inspired by time–memory trade-offs in classical algorithms, we introduce the Linear Reusable Neural Bases (LRNB) architecture, an RNN-based framework that improves parameter efficiency through parameter reuse at the neuron level. Each feedforward residual module is represented as a linear combination of shared neural bases, with depth-specific learnable coefficients and optional shifts providing flexibility across layers. This formulation reduces parameter redundancy across depth and enables the construction of wider and deeper networks within a fixed parameter budget. We further provide a geometric interpretation of the neural bases from a vector-field perspective and extend the framework to linear projection modules. Experiments demonstrate that the LRNB architecture achieves comparable or lower final training loss than independently parameterized baselines while using fewer parameters and maintaining stable training dynamics. These findings support neuron-level reuse as a practical approach to parameter-efficient network design.
Read the full abstract
Machine Learning
NashDreamer: Model-Based Reinforcement Learning for Zero-Sum Imperfect-Information Games
Model-based reinforcement learning (MBRL) has achieved remarkable results in single-agent domains, yet its extension to competitive imperfect information games (IIGs) remains underexplored. In multi-agent settings, opponent-induced non-stationarity complicates the learning process, and decentralized model learning faces severe identifiability barriers, which we argue make centralized model learning a mathematical necessity. Building on this analysis, we propose NashDreamer, a principled MBRL framework for two-player zero-sum IIGs. It introduces a centralized Multi-Agent Recurrent State-Space Model (MARSSM) that decouples environment dynamics from the effect of players' strategies on their individual observations. NashDreamer is designed to use arbitrary policy gradient algorithms and inherits their convergence guarantees towards Nash equilibria under an idealized model. Empirical evaluations across five benchmark games demonstrate that NashDreamer substantially improves sample efficiency over model-free baselines early in the training. Finally, we theoretically analyze the architecture's optimization landscape, identifying the vulnerability of the Dreamer family of algorithms to posterior collapse in stochastic environments. We leave it as an open challenge.
Read the full abstract
Computation and Language
SDARE-Bench: Evaluating Large Language Models on Conversational Stigma Detection and Response in Dyadic and Group Dialogue
Large Language Models (LLMs) are increasingly used in advice seeking and decision making that may affect social judgements. Despite stigma's profound effects on people and communities, benchmarks remain scarce. Existing general-domain evaluations typically rely on static prompts and fixed-format tasks, overlooking conversational contexts and audience effects in everyday communication. To address these gaps, we introduce SDARE-Bench, the first scenario-based benchmark evaluating both stigma detection and open-ended response generation in LLMs, comprising 1,138 dyadic queries and 1,388 group dialogue. Empirical results across 8 LLMs consistently demonstrate poor identification of stigma components, especially in group dialogues. In open-ended response generation, stigma expression was substantially higher in group settings than in dyadic, with weaker resistance to stigma and more unrealistic advice. Responses were evaluated using a classifier trained on 1,392 human annotated responses. In constructed group pressure settings, stigma expression rates further increased to a striking average of 97.5%. Our findings identify stigma response as a recurring LLM safety vulnerability, especially in socially complex conversational contexts.
Read the full abstract
Quantum Physics
Irreducibility and regularisation properties of Gaussian quantum Markov semigroups
We study regularisation and irreducibility properties of Gaussian quantum Markov semigroups (GQMSs) acting on continuous-variable quantum systems. We first identify a natural notion of regularity for operators in this setting, which allows us to formulate and characterise the smoothing effects of GQMSs in terms of algebraic conditions involving the drift and quantum diffusion matrices. These conditions establish a connection with the controllability theory of quantum linear systems and with the structure of decoherence-free subsystems. We then characterise irreducibility through several equivalent algebraic criteria: one formulated in terms of the drift and quantum diffusion matrices, one in terms of the operators appearing in the generalised GKLS representation of the generator, and a third given by a quantum analogue of Hörmander's condition. A central and somewhat surprising consequence is that, in contrast with the classical case, irreducibility is strictly stronger than conditions ensuring regularisation. Our results provide an algebraic framework for analysing these properties and lay the groundwork for a broader study of reducible Gaussian quantum Markov semigroups and, more generally, more general relevant quantum Markov semigroups on continuous-variable systems.
Read the full abstract
High Energy Physics (Experiment)
Construction and Performance of the sMDT Precision Muon Tracking Chambers for ATLAS at the HL-LHC
In order to improve the muon trigger efficiency and the rate capability of the ATLAS muon detectors for operation at the high luminosity upgrade of the Large Hadron Collider (HL-LHC), the Monitored Drift Tube (MDT) tracking chambers in the inner barrel layer of the ATLAS Muon Spectrometer will be replaced by small-diameter Muon Drift Tube (sMDT) chambers integrated with new thin-gap RPC trigger chambers. The sMDT chambers are in serial production since January 2021. This documentation provides insight into the drift tube production and every step performed in the chamber construction at the MPI Munich. The serial production involves a stringent quality control program to assure the reliability and high mechanical precision of the chambers. This program consists of tests of the individual drift tubes, measurements of the chamber geometry and proof of gas tightness. Final certification at the production site is provided using cosmic rays. The dedicated quality control database and monitoring Web interface that is used commonly for both production sites of the sMDT chambers is presented
Read the full abstract
Geometric Topology
Homeomorphisms of surfaces in $4$-manifolds
This paper establishes necessary and sufficient conditions for locally flat knotted surfaces in simply-connected $4$-manifolds to be equivalent. For surfaces with knot group $\mathbb{Z}_d$, we extend results of Lee-Wilczynski from spheres to surfaces of arbitrary genus; the surfaces are permitted to be nonorientable and have boundary. We prove that most projective planes with knot group $\mathbb{Z}_2$ and the same Euler number are determined by the equivariant intersection form of their exterior. We also prove that knots with prime power determinants bound at most one Moebius band in $D^4$ with knot group $\mathbb{Z}_2$ and a given Euler number. Cancellation results lead to new criteria for homologous discs to be equivalent rel. boundary. Finally, we determine the topological extendable mapping class group of knotted surfaces with abelian knot group.
Read the full abstract
Machine Learning (Statistics)
Variable Selection for Feature-Based Newsvendor
Feature-based newsvendor models use observable covariates to tailor inventory decisions, aiming to balance holding and shortage costs under demand uncertainty. However, high-dimensional feature sets often hinder interpretability and inflate data collection and implementation costs. This paper studies variable selection for the feature-based newsvendor problem under a hard cardinality constraint on the number of selected features. We formulate the resulting $\ell_0$-constrained empirical newsvendor problem with $\ell_2$-regularization, establish its computational hardness, and develop a mixed-integer second-order cone programming reformulation that strengthens the standard Big-$M$ formulation. To enable scalability beyond exact optimization, we develop a randomized-rounding algorithm with a bi-criteria guarantee and a greedy heuristic. Statistically, we provide theoretical analysis of the resulting sparse policy estimator, including finite-sample estimation error, out-of-sample risk bounds, and support recovery guarantees. Extensive experiments on both synthetic and real data illustrate the computational and statistical trade-offs among various baselines. Our results demonstrate that the proposed variable selection framework achieves competitive out-of-sample operational costs while using substantially fewer covariates.
Read the full abstract
Signal Processing
Trajectory Analysis of ECG Motif Dynamics in the Run-up to Sudden Cardiac Arrest
Early warning signatures of sudden cardiac arrest (SCA) remain poorly characterised in long-duration ECG. We quantified pre-event changes in ECG morphology using a motif-based trajectory framework. Holter ECGs from 23 patients with annotated SCA were analysed over non-overlapping 10 s windows. Window-level motifs were extracted to quantify trajectories of instability, consistency, dispersion, heterogeneity, and personalised-baseline distance. Each trajectory was normalised to an early baseline using z-scores and aligned to ventricular fibrillation (VF) onset. Abnormal burden was defined as the proportion of windows with z>=3 within a rolling 10-minute window. Median sustained abnormal burden onset occurred 5.8-8.4 h before VF across morphological metrics. Motif dispersion showed the most consistent long-horizon detection, with sustained abnormal burden >=1 h before VF in 100% of patients and >=2 h in 89%. Motif consistency showed the strongest late-stage change, with 70% median abnormal burden in the final 10 min. Peak morphological deviation occurred ~2 h prior to VF. Our label-free framework transforms longitudinal ECG analysis from an event detection approach towards a continuous characterisation of evolving cardiac change, enabling a personalised early warning of SCA, well suited to long-duration wearable ECG monitoring.
Read the full abstract
Audio and Speech Processing
TAG-Bench: Benchmarking Temporal Audio Grounding in Large Audio Language Models
Large audio language models (LALMs) can describe what is heard, but their ability to localize when queried content occurs remains less systematically evaluated. We present TAG-Bench, a benchmark for temporal audio grounding in which a model returns every time interval that matches a natural-language query. TAG-Bench contains 1,750 human-verified query-recording pairs covering 149.5 hours, with eight source-dependent subsets spanning query categories and audio durations from 7 s to 20 min; 22.1% of the queries have multiple ground-truth intervals. Across 21 evaluated systems, the best-performing model achieves 31.2 mIoU and is the only system above 20 mIoU on the two long subsets, yet even this top performer reaches only 21.5% recall at IoU >= 0.7. Moreover, 9 of 21 systems fall below 5 mIoU, and every model under-reports the number of occurrences on one-to-many queries, with none exceeding 13.2% count accuracy. Because responses are free-form, we report parsing-failure rate and MAE coverage: parsing failures remain in mIoU, Recall, gIoU, and count metrics as empty predictions but do not enter MAE. The results separate precise localization, occurrence enumeration, and output-format reliability within a benchmark whose cross-subset comparisons are descriptive rather than controlled estimates of query abstraction or duration. We will release the TAG-Bench data and evaluation code to support future research.
Read the full abstract
Quantum Algebra
Quantum Hopf rigidity in symmetric monoidal categories
We prove that a coaction of a Hopf algebra on another preserving the latter's associative-algebra structure and comultiplication automatically also preserves the counit, antipode, and tensor-interchange map and in fact factors through the coaction dual to an action of an affine group scheme by Hopf-algebra automorphisms. This generalizes and unifies a number of results to the effect that quantum groups have only classical symmetries, due to Kasprzak et al., Budziński-Kasprzak and Brannan et al. The stated automatic classicality follows from the fact that the antipode, counit and tensorand-interchange of a Hopf algebra internal to any symmetric monoidal category are contained in the (non-symmetric) monoidal subcategory generated by the associative-algebra structure and the comultiplication, assuming a weak form of arrow-retraction invariance.
Read the full abstract
Differential Geometry
Stability of the Riemannian positive mass theorem in all dimensions
We prove stability of the Riemannian positive mass theorem in all dimensions, extending the Dong-Song stability theorem. For a sequence of complete asymptotically flat manifolds with nonnegative scalar curvature and ADM masses tending to zero, excising domains whose boundary areas tend to zero yields exterior regions converging to Euclidean space in the pointed measured Gromov-Hausdorff topology. The proof constructs global coordinates from minimal graphs and controls their Hessians using scalar solutions of the conformal Laplace equation on the associated graph metrics.
Read the full abstract
Applied Physics
A Constrained Kuramoto Gradient-Flow System Can Perform High-Accuracy Finite-Time Inference
A central question in physical inference is whether strongly constrained dynamical systems can realize accurate input–output maps through their own finite-time evolution. We study this question in Kuramoto phase networks, whose deterministic dynamics form an input-conditioned gradient flow and whose predictions are read directly from output oscillators. As a constructive training approach, we develop a two-stage teacher–student procedure. A neural teacher is first converted into an explicit phase trajectory whose terminal oscillator activations reproduce the teacher outputs, and the Kuramoto parameters are trained by matching the student vector field along this prescribed path. Because accurate teacher-forced path matching does not ensure accurate autonomous inference, we then differentiate through the autonomous finite-time rollout and directly align its terminal output with the neural target. The resulting oscillator system, with $74$ oscillators, reaches mean test accuracies of $96.711\%$ on MNIST and $86.399\%$ on Fashion-MNIST. This capability persists across neural-teacher architectures, matched system sizes, thermal perturbations, and integration-grid refinement. Together, these results provide a constructive demonstration that a strongly constrained, small-sized Kuramoto gradient-flow system can be trained for high-accuracy finite-time inference through a direct oscillator readout.
Read the full abstract
Optimization and Control
Generalized Semi-Infinite Programming for Robust Optimal Control with Decision-Dependent Uncertainty
Generalized semi-infinite programs (GSIPs) arise in robust optimal control whenever the admissible uncertainty depends on the state or controls. Existing GSIP methods either impose restrictive structural assumptions or require global optimization that scales poorly to control problems. We present a general framework that reformulates any GSIP with mild regularity as an existence-constrained semi-infinite program, smoothing its disjunctive feasibility condition into differentiable existence constraints over a fixed index superset. The resulting program is solved by established adaptive discretization (cutting-plane) methods using only off-the-shelf nonlinear-programming solvers, and converges under standard assumptions. Treating the state trajectory as part of the uncertainty extends the framework to robust nonlinear optimal control with state-dependent uncertainty. We demonstrate it on a nonconvex benchmark GSIP and a satellite de-tumbling problem with dynamically varying inertia.
Read the full abstract
Machine Learning
Quantum Sparse Autoencoders for Q-Matrix Estimation in Cognitive Diagnosis
Q-matrices play a central role in cognitive diagnosis within educational data mining (EDM), specifying which latent skills each assessment item requires. Data-driven Q-matrix estimation remains challenging when assessments involve many correlated skills and when real response patterns depart from idealized generative assumptions. We introduce a novel quantum sparse autoencoder (QSAE) for Q-matrix estimation, which, to the best of our knowledge, is the first application of quantum machine learning (QML) to cognitive diagnosis. Overall, the QSAE embeds each student's binary response vector into a quantum circuit using an encoder, compresses it into a sparse latent representation, and maps that representation to the Q-matrix. We benchmark the QSAE against a classical autoencoder (CAE) across 60 simulated datasets and 9 real-world assessment datasets. The results reveal complementary strengths. Although the CAE partially achieves higher average accuracy under several simulation conditions, the QSAE is substantially more stable across replications, exhibiting lower variance in 49 of the 60 conditions. Moreover, on real assessment data, the QSAE outperforms the CAE on 6 of the 9 datasets. These findings suggest that the principal advancement of QML in this setting is not universal accuracy improvement, but enhanced robustness and capability to explore latent-structure complexity in real datasets.
Read the full abstract
General Relativity and Quantum Cosmology
Bounded Electromagnetic Power Flux at Naked Kasner Singularities
In some occasions, it would appear that classical electromagnetic fields could be able to coexist with strong curvature, naked singularities, in the sense that the electromagnetic power flux would be bounded in any open neighborhood of them. We illustrate this rare behavior by considering the naked singularity present in Kasner spacetimes. There, using the Plebanski–Tamm map between Maxwell theory in curved spacetime and the associated description corresponding to an effective anisotropic medium in flat space, we found not only some regular electrostatic field configurations, but also electromagnetic waves whose amplitudes lead to a well defined, bounded power flux as they approach the singularity. Sometimes, in these unusual situations, the singularity resembles a mirror, perfectly reflecting its environs; in other occasions, the existence of a non-null, finite power flux as the waves travel to or from the singular plane, offers a very simple explanation as to how the singularity might absorb or create radiation fields.
Read the full abstract
Multimedia
Can LLMs Design Video Coding Tools? A Case Study on Planar Mode
This paper explores whether large language models (LLMs) can design video coding tools, a highly challenging task due to the intricate algorithmic coupling of tool modifications. In particular, we present an empirical case study on the Planar mode, a long-standing intra prediction tool in video coding standards. Our experiments operate within a generation-and-evaluation loop, with the LLM generating new Planar predictors, encoder trials evaluating their coding performance, and the LLM re-generating refined implementations based on the evaluation feedback. We first examine directly replacing the default Planar mode in the Fraunhofer Versatile Video Encoder (VVenC) under its faster preset. Experimental results demonstrate that the LLM-generated mode can outperform the conventional Planar mode on this lightweight toolset, achieving 0.18% bitrate savings with 0.4% complexity overhead on the standard benchmark. We further extend our evaluation to the Enhanced Compression Model (ECM). Leveraging newly introduced directional Planar modes, we investigate two integration strategies: directly replacing them, and introducing the LLM-generated predictor as an additional prediction mode with new syntax elements. The empirical results suggest that both strategies can yield coding gains under a constrained low-resolution setting. Overall, this study offers preliminary evidence and practical insights, highlighting both the potential and open challenges of LLM-based coding tool design.
Read the full abstract
Astrophysics of Galaxies
Irreducible-Mass Growth and Extractable Energy in a Self-Gravitating Blandford-Znajek Engine
Wang recently constructed a self-gravitating split-monopole Blandford-Znajek engine and obtained its secular horizon mass and angular-momentum loss rates. We determine the previously unknown consequences of this evolution: the growth of the irreducible mass, the division of rotational work between outgoing electromagnetic power and horizon dissipation, and the lifetime-integrated energy removed from the horizon. For fixed magnetic flux and impedance matching, one half of the instantaneous rotational work is carried outward, and one half irreversibly increases the horizon area. Integrating Wang's spin-down trajectory, we find in the weak-field limit the exact results A_f/A_0=\sqrt{e} and E_{extr}^H/(M_{H,0}c^2)=1-exp{1/4}/\sqrt{2}\simeq 0.09206. The leading maximum-power spin-flux allocation instead gives E_extr^{H}/(M_{H,0}c^2)\simeq 0.04668. These results provide the first irreducible-mass and extractable-energy accounting of Wang's engine. Their extension from quasilocal horizon quantities to energy received at infinity requires a separate flux analysis.
Read the full abstract
Materials Science
No source-free exchange-correlation magnetic fields in non-collinear spin-density functional theory
The source-free condition, $\nabla\cdot\mathbf{B}_{\mathrm{xc}}=0$, where $\mathbf{B}_{\mathrm{xc}}=\delta E_{\mathrm{xc}}/\delta\mathbf{m}$ is the exchange-correlation (xc) magnetic field of non-collinear spin-density functional theory (SDFT) and $\mathbf{m}$ the magnetization, is widely believed to be exact. Recent studies report that projecting the field of a parent functional onto its source-free part improves the predicted magnetic moments. We show that the condition violates the global spin-rotation symmetry of $E_{\mathrm{xc}}$ and therefore cannot be exact. A source-free field is exact only in a transverse theory whose xc functional differs from $E_{\mathrm{xc}}$ and is itself not invariant under spin rotations. We examine a source-free (SoF) construction, an exchange-correlation energy given by a parent SDFT functional evaluated at the transverse magnetization $\mathbf{m}_{\perp}$. Its field is divergence-free, remains a functional derivative and exerts local torques, but still, it breaks global spin-rotation symmetry. With the LSDA as parent we examine the \ce{Mn2} molecule. The SoF and the locally collinear (LoC) LSDA both give too short a bond and too large a bond energy, but SoF finds a broken-symmetry antiferromagnetic solution, consistent with the experimentally observed $^{1}\Sigma^{+}_{g}$ ground state, with a coupling of the right sign and roughly the right size, while LoC finds a high-spin $^{11}\Pi_{u}$ ferromagnet and the wrong sign. The broken symmetry of SoF, however, produces severe errors in directional properties: a rigid spin rotation of the magnetization changes the energy by about $1\,\mathrm{eV}$, in a weak uniform field the magnetization points perpendicular to the field instead of antiparallel to it, and the molecule develops a spurious magnetic anisotropy.
Read the full abstract
Computation and Language
Knowledge Distillation During Mid-Training Favors Reasoning over Factual Recall
Logit-based knowledge distillation (KD) is used to train smaller language models (LMs) via supervision from stronger teachers, but whether its benefits are consistent across training stages remains unclear. Through controlled experiments, we find that forward Kullback-Leibler (KL) distillation–the standard KD formulation–with post-trained teachers behaves fundamentally differently during mid-training, an intermediate phase of self-supervised learning on curated corpora. Surprisingly, while forward KD simultaneously improves reasoning and factual recall during pre-training relative to standard next-token prediction (NTP), it instead slows factual recall acquisition during mid-training despite continued reasoning gains. We trace this stage dependence to an asymmetry in teacher confidence across data domains and the student's evolving knowledge state: teachers are more confident on procedural than knowledge-intensive data, while students acquire low-entropy factual knowledge earlier in training. To mitigate this imbalance, we propose Switch Distillation, a simple mid-training objective that distills on tokens where the teacher is confident, using teacher predictive entropy as a lightweight routing signal, and otherwise falls back to cross-entropy. Switch Distillation consistently outperforms existing distillation objectives across teacher sizes. Relative to standard NTP, it achieves 1.61-1.71x the reasoning performance and 1.13-1.19x the knowledge and commonsense performance while preserving 96.7-96.8% of factual recall. Crucially, these benefits persist after post-training: Switch Distillation closes the factual recall gap while maintaining 1.25-1.32x and 1.13-1.20x gains in reasoning and knowledge and commonsense, respectively.
Read the full abstract
Analysis of PDEs
Failure of Interior Hölder and Gradient Estimates for the Subcritical Special Lagrangian Equation
For every $n\ge3$, we show, by a family of explicit entire real-analytic solutions, that, for every subcritical phase and every $0<\alpha\le1$, oscillation-based interior $C^\alpha$ estimates fail for the special Lagrangian equation in $\mathbb{R}^n$. In fact, this family has no common interior modulus of continuity.
Read the full abstract
Computer Vision
Revisiting Cross-View Completion: Self-Supervised Pre-Training via Reconstruction Error Comparison
Self-supervised pre-training via cross-view completion learns strong features for 3D vision from co-visible regions of image pairs. However, the reference view provides little information for reconstructing non-co-visible patches, implicitly yielding a monocular training signal in these regions. We introduce Gekko, which turns this limitation into a useful signal. The relative improvement of the cross-view reconstruction error over a masked-autoencoder error is a self-supervised proxy for co-visibility: large improvements indicate co-visible regions, negligible ones non-co-visible areas. Gekko is a network, trained from scratch, that jointly performs cross-view completion, masked autoencoding, and per-pixel prediction of this relative improvement, providing an additional binocular signal for all masked regions without any ground-truth 3D annotation. Under identical architectures and training data, Gekko consistently outperforms CroCo on zero-shot correspondence estimation, relative pose estimation, and pointmap regression, with up to 6 times higher accuracy at the strictest relative-pose threshold and a 22% drop in end-point error on ETH3D. The extra channel it learns is itself a strong co-visibility detector on unseen scenes, and Gekko's frozen features outperform released cross-view backbones of comparable or larger size. It can also be trained directly from raw videos with a simple stride-based curriculum, removing the cumbersome 3D preprocessing prior methods require while matching models trained on curated data. Code and pre-trained models are publicly available.
Read the full abstract
Probability
Multiplicative comparisons of Rényi entropies for weighted Bernoulli sums
We establish multiplicative comparisons between Rényi entropies of different orders for weighted sums of independent Bernoulli random variables. In particular, we prove a logarithmic comparison between the zeroth- and infinity-order Rényi entropies, which yields a polynomial improvement over the square-root bound of Jain, Sah, and Sawhney. As an application, this leads to an improved parameterized running time for the randomized bin-packing algorithm of Nederlof, Pawlewicz, Swennenhuis, and Wȩgrzycki. We also obtain explicit dimension-free, constant-factor comparisons between Rényi entropies of positive orders.
Read the full abstract
Computational Geometry
Sierpiński–Knopp Wasserstein Distance for Persistence Diagrams and Applications to 2-Wasserstein Approximation
This paper introduces the Sierpiński-Knopp (SK) Wasserstein distance, a fast metric between persistence diagrams. The SK-Wasserstein distance, denoted $d_{\mathrm{SK}}$, maps diagram points and their diagonal projections to the unit interval via the Sierpiński-Knopp space-filling curve on the upper diagonal triangle. The encoded point sets are then efficiently matched via one-dimensional optimal assignment, in \(O(N\log N)\) steps, yielding an explicit diagonal-aware point assignment between the two input persistence diagrams. We show that the SK-Wasserstein distance controls the classical \(2\)-Wasserstein distance between diagrams, admits an explicit isometric embedding into a Hilbert space, and induces a positive-definite Gaussian kernel, making the resulting geometry directly compatible with Euclidean and kernel-based learning methods. A tighter surrogate dissimilarity, noted \(W_\Gamma\), is also introduced based on the point assignments along the curve. Experiments on 12 scientific collections comprising 227 diagrams show median per-collection speedup of \(d_{\mathrm{SK}}\) over state-of-the-art approximations of \(W_2\) is \(626\times\), while the aggregate speedup over the full benchmark is \(2100\times\). Average-linkage partitions obtained from \(d_{\mathrm{SK}}\) and \(W_\Gamma\) each exactly match the corresponding \(W_2\) partition on 8 of the 12 collections. Hilbert \(k\)-means and Gaussian spectral clustering, both based on \(d_{\mathrm{SK}}\), achieve mean adjusted Rand indices (ARI) of \(0.756\) and \(0.800\), respectively, with respect to the benchmark reference partitions, compared to \(0.750\) obtained by average linkage on \(W_2\). The Gaussian \(d_{\mathrm{SK}}\) kernel supports other kernel-based analysis tasks, as illustrated by its use for contiguous segmentation of ordered diagram collections in our experiments.
Read the full abstract
Hardware Architecture
Performance Characterization of SPEC CPU 2026 on AMD EPYC 9755 Processor
SPEC CPU 2026 is the first major update to the industry-standard CPU benchmark suite since 2017. This paper presents the first microarchitecture based performance characterization of the new suite, conducted on AMD EPYC "Zen 5", also the first SPEC CPU characterization study on this microarchitecture. Using a multi-lens methodology spanning pipeline efficiency, control flow behavior, cache hierarchy pressure, and instruction mix, we analyze both SPECrate and SPECspeed suites. We introduce scale analysis, comparing single-copy to full-system behavior to expose system-level bottlenecks invisible to conventional characterization. Our analysis reveals substantial behavioral diversity across the suite, and the multi-lens analysis identifies three distinct behavioral clusters: frontend control-flow-dominated workloads that stress branch predictor throughput rather than accuracy, high-efficiency compute workloads that suffer SMT contention at scale, and memory bandwidth-bound workloads with poor L3 filtering even at single-copy. Scale-dependent effects, including SMT dispatch contention causing throughput reduction and L3 capacity interference, emerge only at full system utilization. This work establishes an empirical foundation for architectural research and workload-driven design decisions targeting next-generation datacenter processors.
Read the full abstract
Artificial Intelligence
EvoSCM: Scientific Belief Revision Through Causal Model Evolution and Experimentation
Scientific discovery depends on the ability to form hypotheses, test them through experiments, and revise them when evidence disagrees. Existing LLM agents support this process by improving their reasoning or actions, but their scientific beliefs are often scattered across free-form reasoning and difficult to update coherently. This makes it difficult to identify what failed, what should change, and whether revisions remain consistent with prior evidence. We introduce EvoSCM, which represents scientific beliefs as a population of structural causal model (SCM) hypotheses that can be tested and revised across experiments. EvoSCM formulates scientific discovery as a closed loop in which causal hypotheses guide experimentation and experimental outcomes drive causal model evolution. Competing SCM hypotheses make falsifiable predictions and guide discriminative experiments that separate alternative explanations. When observations contradict these predictions, EvoSCM distills discrepancies into correction rules identifying which aspects of the hypotheses fail to explain the evidence. These rules guide revisions to causal dependencies, latent factors, mechanisms, and parameters. Revised hypotheses are validated against accumulated evidence and carried forward to guide subsequent experiments, allowing scientific beliefs to evolve cumulatively. We evaluate EvoSCM across physics, chemistry and materials, and biology. It consistently outperforms baseline agents and existing evolution methods, yielding more accurate explanations and predictions with more effective use of experimental budgets. The evolved SCMs also transfer across base models, suggesting reusable scientific knowledge beyond any single model's reasoning process.
Read the full abstract
Databases
Relational-Core Graph Analytics Querying graphs at SQL scale, and why the node/edge model is a performance tax, not a truer picture of connected data
A durable assumption holds that graph analytics requires a purpose-built graph engine, and that relational systems are ill-suited to connected data. We argue the opposite for the workloads enterprises actually run. A columnar relational engine fronted by a graph query language matches or exceeds native graph engines on analytical graph queries, and - decisively - scales past the point where in-memory graph engines fail. We further argue that the node/edge property graph is not a more faithful model of connected data but a re-encoding of relationships that already exist explicitly in relational tables; reconstructing them at query time is pure overhead. We present ClickGraph and its Databricks-dialect sibling DeltaGraph, systems that translate Cypher directly onto the native relational schema - the tables, columns, and foreign keys as they already exist - and execute in place on ClickHouse, Databricks, or in-process on lakehouse files, with no import and no separate cluster. Because the output is ordinary SQL, an underperforming query is an open optimization surface: it can be rewritten, and the engine itself extended. We support the argument with a peer system's own published benchmark, in which a columnar engine outruns Neo4j by two-to-four orders of magnitude, and with reproducible measurements across the LDBC Social Network Benchmark suite.
Read the full abstract
Quantum Physics
Quantum Weighted Moving Average for Predicting Limit Order Book Trends
Can quantum computers be useful for forecasting multivariate financial time series? In this work, we consider the problem of predicting price trends from limit order book (LOB) data. After identifying key components of classical models, we introduce the quantum weighted moving average (QWMA) model. The two main building blocks are, first, classically preprocessing via normalization of both feature and temporal dimensions and, second, a linear combination of unitaries-based layer for unitarily-embedded classical data. The models are evaluated on the FI-2010 benchmark dataset and a second dataset of China A-share stocks. While we do not present evidence of quantum advantage, the combined classical-quantum model demonstrates performances close to the best classical models. The quantum part of the model is expressive enough to focus on the most predictive parts of the time series. Several specializations of the QWMA model are considered, in particular a variant related to the widely-used exponential moving average (EMA). We give consideration to the limitations of the methods and the datasets.