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

Friday 11 September

Quantum Physics

Spatial symmetry in few-photon interference and quantum metrology

In previous work [Physical Review Letters, 136(6):060807, 2026], we developed a symmetry-based framework for generalized Hong-Ou-Mandel interference and its applications to quantum metrology. Here, we apply this framework to experimentally accessible configurations involving a few photons and explore how it sheds new light on a variety of existing results while suggesting new experimental perspectives. We first reinterpret a two-photon Mach-Zehnder interferometer as a generalized Hong-Ou-Mandel experiment and show how its photon-counting statistics and metrological performance follow from the spatial symmetry of an effective probe state. We then study two coherently pumped parametric photon-pair sources, derive their complete output distribution, and connect source indistinguishability with spatial symmetry. Finally, we extend the discussion to a three-mode interferometer and show that the same symmetry principles provide a natural description of three-photon interference in tritter configurations.
Read the full abstract

Machine Learning

Convergence of Stochastic Gradient Methods under Heavy-Tailed Noise and Hölder Smoothness

Classical convergence guarantees for stochastic gradient methods typically assume Lipschitz-smooth objectives and finite-variance gradient noise, both frequently violated in practice. In contrast, we study nonconvex stochastic optimization under the joint relaxation of these assumptions: objectives with $(L,s)$-Hölder continuous gradients, $s\in(0,1]$, and gradient noise satisfying only a bounded $\alpha$-th moment condition for $\alpha\in(1,2]$. We establish three convergence results. Firstly, that standard SGD converges at rate $O(T^{-s/(1+s)})$ whenever $\alpha\ge1+s$, extending the classical nonconvex SGD rate to heavy-tailed noise and Hölder smoothness simultaneously. Secondly, we analyze $\delta$-regularized gradient clipping ($\delta$-GClip), a provable trainer of wide and deep nets, and establish a stationarity rate of $O(T^{-2s(\alpha-1)/[(1+s)(2\alpha-1)]})$ under the same condition. Thirdly, we analyze standard gradient clipping (G-Clip) and show that it recovers the above rate for $\alpha\ge1+s$ while in the very heavy-tailed regime $\alpha<1+s$, it has a convergence rate $O(T^{-2s(\alpha-1)/[(\alpha-1)+s(2\alpha-1)]})$ — the first convergence guarantee in this regime for any stochastic gradient based method.
Read the full abstract

General Relativity and Quantum Cosmology

Compact binary systems in the post-Newtonian limit of gravitational theories with preferred frames

Gravitational-wave observations of compact-binary inspirals offer the opportunity to test deviations from general relativity in both strong and weak-field regimes. In previous work, we developed a theory-independent formalism for describing compact-binary dynamics in metric theories of gravity without preferred frames, using the parametrised post-Newtonian (PPN) couplings together with an additional set of scalar sensitivity parameters that encode violations of the strong equivalence principle. Here, we extend this framework to include preferred-frame effects, of the type that can arise in (for example) vector-tensor theories of gravity. This is achieved by introducing a time-like vector field that identifies the preferred frame, and then by allowing compact body masses to depend on the invariant formed from this field and the body's four-velocity. We calculate the modified equation of motion up to first post-Newtonian order (1PN), where the preferred-frame PPN parameters $\alpha_1$ and $\alpha_2$ acquire contributions from our new vector sensitivities. We further combine scalar and vector sensitivities in a unified formalism, which we validate with an example scalar-vector-tensor theory. This approach provides the orbital dynamics required to extend gravitational-wave constraints on PPN and sensitivity parameters to theories of gravity that admit preferred frames, as well as those that do not.
Read the full abstract

Networking

Vertical Assessment of RF-EMF Exposure in a Building Adjacent to a Multi-Operator Shared Base Station

This paper extends a previously published proposed approach for estimating worst-case exposure scenarios in buildings using publicly available Base Station (BS) parameters, applying it to a critical case involving a multi-operator shared site. Radiofrequency Electromagnetic Fields (RF-EMF) exposure was assessed on every floor of a 20-story building facing the site, revealing significantly higher levels near the center of the antennas' estimated vertical zone of interest, with a peak of 31.86 V/m, 23.95 times higher than the street-level peak value. However, due to uncertainties in the input parameters, measurements across all locations within the zone of interest are required to accurately identify the floor of maximum exposure. The results demonstrate the usefulness of the proposed estimation approach and highlight the need for improved methods that consider BS antenna configurations and current technical information to ensure accurate assessments of exposure to RF-EMF.
Read the full abstract

All authors Ricardo Q. de F. H. Silva, Marcio E. C. Rodrigues, Fred S. R. Pinheiro, Gutembergue S. da Silva, Halysson B. Mendonça, Vicente A. de Sousa Jr

Dynamical Systems

The Hausdorff dimension of the spectrum of the Thue-Morse Hamiltonian

In 2015, Liu, Qu (Commun. Math. Phys., Vol. 338) show that the Hausdorff dimension of the spectrum of the Thue-Morse Hamiltonian admits a positive lower bound independent of the coupling constant lambda: dim_H sigma(H_{tm,lambda}) ge (log 2) / (140log 2.1). In this paper, we prove that the spectrum of the Thue–Morse Hamiltonian has full Hausdorff dimension for any lambda in mathbb{R}: dim_H sigma(H_{tm,lambda})=1.
Read the full abstract

High Energy Physics (Theory)

Accelerating black holes in higher-derivative gravity

Accelerating black holes in General Relativity are described by the C-metric. We study how this geometry is modified by higher-derivative corrections, focusing on cubic curvature terms of even and odd parity. Using a suitable metric ansatz, we obtain closed-form solutions at first order in the coupling constants. The corrected geometries contain new integration constants, which correspond to physical deformations of the C-metric, and exhibit different behavior depending on the parity of the correction. In particular, parity-violating terms generate an off-diagonal metric component and induce rotation. For AdS solutions, we analyze the thermodynamics, including the Hawking temperature, entropy, and string tensions, and discuss the subtleties in defining the mass and thermodynamic string lengths.
Read the full abstract

Number Theory

Tight universality of $m$-gonal forms with minimal criterion sets

For integers $m\geq3$ and $n\geq1$, an $m$-gonal form is called tight $\mathcal{T}(n)$-universal if it represents exactly the positive integers $\mathcal{T}(n)=\{ n, n+1, n+2, \ldots \}$. In this paper, we study the minimal criterion set $\mathrm{CS}(m,n)$ for tight $\mathcal{T}(n)$-universality. Our main result determines $\mathrm{CS}(m,n)$ for all $n \geq8$, $3 \leq m \leq \left\lfloor \frac{3n+1}{2} \right\rfloor$. More precisely, \[ \mathrm{CS}(m,n)= \begin{cases} \{ n, n+1, \ldots, 2n-1 \}, & m=5,\newline \{ n, n+1, \ldots, 2n \}, & m\neq5. \end{cases} \] We also establish the corresponding tight $\mathcal{T}(n)$-universality results and show that the upper bound on $m$ is optimal.
Read the full abstract

Numerical Analysis

Total Generalized Variation for Inverse Problems Involving Piecewise Linear Finite Elements

Higher-order regularization has shown improved results over first-order methods such as total variation. In this paper, convergence of a piecewise linear finite element discretization is shown for the second-order total generalized variation functional. Additionally, this convergence result is extended to a spatiotemporal setting after the spatiotemporal function space is comprehensively introduced. In both cases, discrete minimizers converge to the continuous minimizer with rate $h^{1/4}$, matching the best known rate for finite element discretizations of total variation in the same general setting. Numerical experiments confirm the convergence rates and indicate faster convergence in practice. Comparisons with related spatial image reconstruction algorithms show comparable reconstruction results. The spatiotemporal case is discussed using the example of the inverse problem in electrocardiographic imaging, where the proposed functional improves the state of the art.
Read the full abstract

All authors Moritz Kappes, Manuel Haas, Thomas Beiert, Simone Pezzuto, Alexander Effland

Quantum Physics

Average-case hardness of Betti number estimation

We establish the average-case hardness of Betti number estimation on random clique complexes via a reduction from the planted clique problem. We further show that our reduction implies a series of hardness results for many problems in both classical and quantum Topological Data Analysis (qTDA). Under the classical planted clique conjecture, no randomized polynomial-time Betti number estimator achieves additive error below $\tfrac12$ with constant advantage. Under a new quantum planted clique conjecture that we introduce, the same conclusion holds for quantum polynomial-time algorithms. We also obtain related conditional hardness results for homology vanishing, additive approximations with larger error tolerances, preparation of simplex and harmonic states, cycle recovery, and counting eigenvalues at low energy. Our reduction clarifies the structural requirements for quantum advantage in TDA and provides a new lens to investigate the classical and quantum complexity of related problems.
Read the full abstract

High Energy Physics (Experiment)

The DALI Haloscope: A Magnetized Phased Array Coupled to a Semi-Open Fabry–Pérot Resonator

DALI is an axion haloscope consisting of a magnetized phased array backed by a conducting mirror and coupled in the near field to a semi-open Fabry–Pérot resonator. The applied static magnetic field makes both the dielectric interfaces and the conducting mirror sensitive to axion-induced electromagnetic conversion. Each interface acts as a radiating surface, and the resulting fields are projected onto the mode collected by the resonator and delivered to the receiver. We derive the detected power in terms of the power available from the magnetized mirror, the coherent contribution of the dielectric interfaces, the transverse mode overlap, the loaded quality factor, and the receiver-coupling coefficient. This formulation separates two distinct enhancement mechanisms. The dielectric-interface emissions may add coherently, producing a dielectric boost, while the semi-open Fabry–Pérot resonator provides enhancement through resonant storage of the coupled electromagnetic field. These contributions depend differently on the dielectric thicknesses, spacings, and resonant mode, and should therefore be evaluated independently. The same assembly can operate between the limits of a coherently boosted dielectric haloscope and a mirror-sourced, resonantly enhanced Fabry–Pérot haloscope. This combined architecture offers a flexible approach to resonant axion searches at frequencies for which conventional closed microwave cavities become increasingly limited in conversion volume.
Read the full abstract

All authors Javier De Miguel, Antonios Gardikiotis, Eduardo David González-Carretero, Elvio Hernández-Suárez, Roger J. Hoyland, Enrique Joven, Abaz Kryemadhi, Marios Maroudas, Chiko Otani, J. Alberto Rubiño-Martín, Yannis K. Semertzidis, Michael E. Tobar, Konstantin Zioutas

High Energy Physics (Experiment)

Constraints on Galactic Dark Photons from a DALI Prototype

An analysis of 36 hours of data from a DALI haloscope reveals no statistically significant excess attributable to dark-photon dark matter. We therefore set new constraints in the 6.88691–6.91792 GHz band, reaching a dark-photon-to-photon kinetic mixing strength of $\chi\lesssim 6.8\times10^{-14}$ at $28.54 \, \mu\mathrm{eV}$. To our knowledge, this result established the strongest laboratory-based exclusion limit in this frequency range.
Read the full abstract

All authors Javier De Miguel, Antonios Gardikiotis, Elvio Hernández-Suárez, Roger J. Hoyland, Enrique Joven, Abaz Kryemadhi, Haroldo Lorenzo-Hernández, Marios Maroudas, Chiko Otani, J. Alberto Rubiño-Martín, Yannis K. Semertzidis, Michael E. Tobar, Konstantin Zioutas

Sound

DiffSynth-Music: Audio-Conditioned KV-Cache Adapters for Controllable Music Generation

Text and lyrics specify broad musical characteristics and sung content but offer limited control over musical timing, melody, and reference-based style. We introduce DiffSynth-Music (https://modelscope.cn/models/DiffSynth-Studio/DiffSynth-Music), a framework that adds composable audio conditioning to a music synthesis backbone through layer-wise key-value injection. The three template models, Control, Prosody, and Reference, are initialized from the backbone diffusion transformer and trained with conditional flow matching. They support five control types: beats, vocals, accompaniment, prosody, and reference audio. A shared variational autoencoder maps conditioning waveforms into a common latent space, enabling their attention memories to be combined. With the template timestep fixed at the clean-data endpoint and other inputs held constant, each control cache is computed once and reused throughout sampling. Training pairs are derived from music recordings using beat extraction, source separation, vocal resynthesis, and reference-excerpt selection. Single-control evaluations on Mandarin and English songs demonstrate improved adherence across all five control types and better lyric fidelity under vocal conditioning relative to the backbone. Automatic music-quality and instruction-following scores remain broadly comparable to those of the evaluated base models, with metric-specific trade-offs. We release the three template models to support research and creative applications in controllable music generation.
Read the full abstract

All authors Zhongjie Duan, Shengchuan Gao, Hong Zhang, Yingda Chen

General Relativity and Quantum Cosmology

Intrinsic pressure anisotropy in spherical Proca stars

Pressure anisotropy in relativistic stars is commonly prescribed through a phenomenological closure, obscuring its microscopic origin and relation to stress-energy conservation. Here it is derived directly from the minimally coupled Einstein-complex-Proca theory. For spherical Proca stars, the principal-pressure difference admits an exact on-shell form whose sign is controlled solely by the local mass-shell threshold. The stress is radially dominated in the core, reverses on a surface of fixed gravitational redshift, and becomes tangentially dominated in the envelope. At the first mass maximum, the fractional anisotropy reaches about 21% near the density maximum, whilst the region beyond the reversal contains about 9% of the mass. Its exact atmospheric limit is approximately 24%, equal in magnitude and opposite in sign to the scalar-boson-star limit. Because the usual local fluid variables remain non-zero at the crossing, no sign-definite closure constructed from them can reproduce the profile. These results identify an intrinsically vectorial stress reversal, generated without additional interactions, and provide a first-principles benchmark for anisotropic bosonic compact objects.
Read the full abstract

Optics

Interplay between growth conditions and formation of localized light emitters in epitaxial AlN thin films

AlN is a promising material for integrated quantum photonics due to its broad transparency window, its compatibility with CMOS technology, and the possibility to host localized light emitters acting as single photon sources. However, little is understood yet about the microscopic nature and formation mechanisms of these light emitters. In this contribution, we demonstrate that their formation efficiency depends significantly on the growth conditions of the AlN thin film. We reveal that localized light emitters with distinct zero-phonon lines in the 600–800~nm range can be efficiently formed in epitaxial AlN thin films grown under N-rich conditions by exposing the AlN to a post-growth treatment consisting in a proton irradiation followed by an annealing process. In contrast, the density of light emitters formed in AlN grown under Al-rich conditions remains low, even after the thin films have been exposed to the post-growth treatment. These results open a pathway toward a better understanding of the microscopic nature of these light emission centers, as well as their engineered integration as quantum light sources in AlN-based optomechanical platforms.
Read the full abstract

All authors Meysam Saeedi, Duc V. Dinh, Andrey N. Anisimov, Mingyun Yuan, Georgy V. Astakhov, Alberto Hernández-Mínguez

Artificial Intelligence

MPT: Missing Prototype Tracking via Barycentric Reconstruction in Vehicular Federated Learning

Cross-vehicle federated learning enables vehicles to collaboratively improve perception models while keeping locally collected driving data private. However, vehicle participation is transient, and a vehicle may depart before training converges while permanently taking its local data. When this departing vehicle holds most samples of a target class, the class becomes rare in the remaining FL network, and its recognition can silently degrade as the shared backbone continues to evolve. Recovering the class is difficult since the few remaining samples provide a noisy prototype estimate, while FL privacy constraints prevent centralized access to raw data or per-sample features. This paper presents MPT, a cross-vehicle FL framework that maintains rare-class recognition by reconstructing its prototype at every round from privacy-preserving class-level statistics. MPT combines a barycentric decomposition that tracks drift shared with remaining-class prototypes, a covariance-based residual prediction that estimates out-of-span drift, and an adaptive calibration that weighs the remaining rare-class samples according to their reliability. We evaluate MPT on three vehicle classification tasks and four backbones against representative calibration and drift-compensation baselines. MPT outperforms all baselines in rare class F1, reaching 0.516 on the nuImages dataset with only 1% of rare-class samples remaining, without raw data, per-sample features, or retraining.
Read the full abstract

All authors Hanju Jang, Gyeongmin Han, Sungmin Lee, Kichang Lee, Chunghan Lee, JeongGil Ko

Software Engineering

Detecting HTTP Status Code Misuses in REST APIs via Static and Dynamic Analysis

REST APIs are widely used on the web for client-server communications. As REST is based on HTTP, server responses contain status codes to indicate the outcome of requests (e.g., 200 OK for a success and 404 Not Found for an unavailable resource). While HTTP status codes are standardized, their semantics are not enforced in REST, leading to many misuses in practice (e.g., using 500 Internal Server Error to describe a client error). Such misuses may have nefarious consequences, such as reducing interoperability, misleading API clients, or causing false positives in testing tools. In this paper, we present a combined static and dynamic analysis approach for detecting HTTP status code misuses in REST APIs. We first study 2,625 real-world REST API specifications to identify relevant status codes and derive a set of 30 usage rules based on HTTP standards and REST API principles. We then implement tools to identify such rule violations in OpenAPI specifications (static analysis) and in API behavior (dynamic analysis). Our evaluation finds that status code misuses are frequent and systematic in REST APIs, with both static and dynamic approaches detecting various misuses. We highlight that both approaches may be used in a complementary manner, and also provide insight for REST API testers and users alike.
Read the full abstract

All authors Alix Decrop, Andrea Arcuri, Mike Papadakis, Pierre-Yves Schobbens, Gilles Perrouin

Artificial Intelligence

Unified Agentic Video Editing Across Levels of Complexity and Creativity

Editing is a core component of video production, requiring creative planning and decisions under multiple constraints. Here, we report methods for agentic tooling for automated video editing across three tasks varying in editorial goal, complexity and creativity, namely scene previews, video summaries and cinematic trailers. We evaluate the outputs and discuss implications for automation and agency.
Read the full abstract

All authors Surabhi S. Nath, Kim Ferres, Milan Petrović, Lion Schulz

Optimization and Control

Infinite Horizon Mean-Field Terminal Value Problem

We introduce a class of finite- and infinite-horizon discrete-time control problems for studying macroscale systems in discrete time, which we refer to as mean-field terminal value problems (MFTVPs). An MFTVP takes a terminal \(Q\)-function as input and, starting from this terminal \(Q\)-function, seeks an indefinite backward evolution of optimal policies and state measures, as required for an infinite-horizon non-stationary mean-field equilibrium (MFE) of a discrete-time mean-field game (MFG). We study the existence and uniqueness properties of this problem. As an application, using the uniqueness properties of MFTVPs, we show that, in the average-cost setting (and in the discounted setting for sufficiently small discount factors), MFTVPs can be used as an intermediate step to develop approximation schemes between finite-horizon MFEs and average-cost MFEs (respectively, stationary MFEs). In contrast to approaches that use infinite-horizon MFGs as intermediate problems, MFTVPs can be employed when the discount factor is sufficiently small (or, in the average-cost setting, when the drift factor is sufficiently small), together with a regularizer, to approximate a stationary MFE. Assuming uniqueness of the solution of a MFTVP, the restrictions we impose on the Lipschitz parameters of the system are weaker than the contraction conditions available for regularized stationary MFGs, and our approximation scheme remains applicable even in the presence of multiple stationary MFEs.
Read the full abstract

Materials Science

Crack propagation threshold in single-crystal silicon: a cleavage plane-crackons model

Experiments revealed the discontinuities in the dependence of the crack speeds on the crack driving force. Despite great efforts, until now, previous theoretical methods, such as linear elastic fracture mechanics and molecular dynamics calculations, failed to elucidate the issue of speed gap. Herein, the cleavage plane-crackons model for crack propagation has been proposed. The normalized crack propagation speed of silicon is proportional to the one-fourth power of the quantum numbers. This motion equation is consistent with experimental results of silicon. It clarifies the underlying mechanism of the long-standing issue of the speed gap. The relationship between quantum numbers and the roughness of the surface of the cracks has been established. The critical quantum number for the onset of unstable crack propagation in single-crystal silicon has been determined. The correspondence between the critical quantum number and fracture toughness has been discovered. This methodology lead to insights into the underlying mechanism of the fracture processes. It is not limited to silicon and can be extended to other crystalline material to understand and predict the fracture behaviors.
Read the full abstract

Information Retrieval

A Historical Corpus Is Not a Historical System: Auditing Hindsight Leakage in Stateful Data Discovery

Offline replay should estimate what a discovery system could retrieve at a historical point, yet freezing the corpus leaves interaction memory unconstrained. We formalize point-in-time (PIT) discovery through historical state $(D_t, \theta_t, M_{< i})$ and introduce a paired replay that changes only memory availability. The protocol constructs PIT and full-stream Future views from behavior-only traces and audits selected entries with a Temporal Violation Rate. Across three table-text domains, two stream regimes, two retrievers, and five seeds (216,000 rows), Future inflated Asset Recall@100 by 2.62-5.24 points; all 12 paired intervals excluded zero. With behavior-only trace memory, PIT underperformed the no-memory Stateless condition; Future masked 32.7-48.4% of that harm. For a simulated positive-feedback cache, PIT added 4.65-18.96 points over Stateless while Future added another 4.11-9.58 points. On five timestamped FreshStack topics, Future exceeded PIT by 2.72 points [1.75, 3.71]. Historical evaluation must version and validate memory with the corpus.
Read the full abstract

All authors Yixi Zhou, Fan Zhang, Sikun Wang, Yingfan Xu, Haipeng Zhang

Optimization and Control

High-Probability Convergence of SGD via Batched Updates

Stochastic gradient descent (SGD) is the primary workhorse for large-scale optimization. While the average behavior of its iterates, typically characterized by mean-squared error bounds, is well-understood, obtaining high-probability guarantees for the last iterate remains challenging. Prior approaches to this problem have either imposed restrictive assumptions (such as bounded domains or gradients) or relied on complex proofs involving auxiliary sequences. In this work, we propose Batched SGD, a simple variant that partitions online samples into epochs and performs a single update per epoch using a refined, low-variance gradient estimate. Our main contribution demonstrates that this batching mechanism enables a surprisingly simple high-probability analysis that avoids both restrictive assumptions and auxiliary sequences. Under standard smoothness and norm-sub-Gaussian noise assumptions, we establish near-optimal rates for both strongly convex and non-convex objectives. Furthermore, we show that our batching idea extends naturally to federated learning (FL). We provide the first high-probability guarantees for FL, achieving logarithmic communication complexity, linear speedup in the number of agents, and resilience to data heterogeneity.
Read the full abstract

Number Theory

The first moment of quadratic Dirichlet $L$-functions in the even hyperelliptic ensemble

We establish an asymptotic formula for the first moment of the central values of quadratic Dirichlet $L$-functions in the even hyperelliptic ensemble $\mathcal{H}_{2g+2}$ over $\mathbb{F}_q[x]$, for every fixed odd prime power $q$. Working with the completed $L$-function yields an exact two-term central-value formula with truncation levels $g$ and $g-1$; combining the two truncations before the square-dual contour shifts leaves the three cubic points $z^3 = q^{-4}$ as the only secondary poles, and they contribute the secondary term $q^{2g/3+2}(a_g g + b_g)$ with an error $O_\varepsilon(q^{g(1+\varepsilon)/2})$. The coefficients are real and $m \mapsto (a_m, b_m)$ has minimal period exactly three, so the coefficient of $g$ at the order $q^{2g/3}$ depends on $g$ modulo 3, in contrast with the earlier even-degree formulas; the difference is confirmed numerically.
Read the full abstract

Optimization and Control

Sharp Linear Bounds and Exponential Separation for Single-Flip Methods in Absolute Value Equations

Rohn's sign-accord algorithm is a classical single-coordinate sign-flip method for $Ax-|x|=b$, selecting the least-index mismatched coordinate at each step. We construct rational instances on which this rule requires exactly $2^{n-1}-1$ flips from natural initialization, even when both $\|A^{-1}\|_1$ and $\|A^{-1}\|_\infty$ are arbitrarily small. We propose two alternative selection rules, Gauss–Southwell and steepest single flipping, which require only one flip on the same instances. A discrete potential framework with positive polarity factors identifies solutions with global maximizers. Under positive polarity with ratio $\kappa$ and $r=\|A^{-1}\|_\infty<1$, both rules terminate from any initial signature after at most $\lfloor s\kappa(1+r)/(1-r)\rfloor$ flips, where $s$ is the number of nonzero solution coordinates. For fixed $\eta_1<1/2$ and $\eta_2<1$, both rules have worst-case flip complexity $O(n)$ on the class $\|A^{-1}\|_1\le\eta_1$ and $\|A^{-1}\|_\infty\le\eta_2$. An $n-1$ lower bound from natural initialization shows that this linear order is sharp. Under stronger, explicit sign-correction conditions, each rule makes exactly as many flips as there are initially incorrect signs among the nonzero solution components. Thus, each rule makes at most \(n\) flips, and this bound is optimal. A five-dimensional example requiring six greedy flips shows that the bound of $n$ does not extend to the entire common small-norm class. Building on Radons–Rump sign certificates, we further develop three classes of sign-flip algorithms using elimination guidance, cyclic scanning, and randomized $m$-flipping. Under $\rho(|A^{-1}|)<1/2$, we prove at most $n$ flips for elimination guidance, $n^2$ inspections for cyclic scanning, and $O(n^2/m)$ expected batch updates for randomized $m$-flipping. Exact computations and retained paired experiments distinguish flip counts from computational cost.
Read the full abstract

Fluid Dynamics

Gas-liquid stratified MHD flows in inclined rectangular ducts

This study investigates fully developed stratified gas/liquid magnetohydrodynamic (MHD) flow of an electrically conducting liquid and a nonconducting gas in inclined rectangular ducts subjected to a vertical magnetic field. Analytical and numerical solutions for the velocity and induced magnetic fields are obtained in terms of the governing dimensionless parameters for concurrent upward, concurrent downward, and countercurrent flows. Unlike single-phase MHD flow, duct inclination strongly affects two-phase flow by altering the liquid holdup and the relative contributions of gravitational, frictional, and electromagnetic forces. The results reveal a complex interplay among gravity, Lorentz forces, and wall and interfacial shear stresses. These interactions govern the liquid holdup, pressure gradient, multiple steady solutions, flooding limits, local backflow, jet-like velocity structures, and pumping requirements. Wall conductivity critically affects the induced magnetic field and Lorentz force distribution and therefore cannot be neglected, even at very small magnetic Reynolds numbers. Fully insulating ducts generally exhibit the weakest electromagnetic effects and behavior closest to non-MHD flow. Configurations with a conducting bottom wall exhibit substantially stronger electromagnetic effects and greater sensitivity to side-wall conductivity, leading to pronounced changes in the velocity field, liquid holdup, pressure gradient, gas-lubrication effect, and overall pumping-power requirements.
Read the full abstract

All authors Subham Pal, Ilya Barmak, Arseniy Parfenov, Alexander Gelfgat, Neima Brauner

Computer Vision

Same Encoder, Different Winner: A Paired-View Framework for Cell Painting Encoder Evaluation

Vision encoders for Cell Painting are typically ranked by a single evaluation, commonly replicate mean average precision (mAP). We introduce CP-BG-Bench, a paired-view evaluation framework that holds the central cell fixed across four matched views (raw crop C, segmented S, and density-augmented variants CD and SD), ablating or augmenting surrounding pixels as a controlled intervention. Instantiating the framework on three datasets (JUMP-CP, RxRx1, RxRx3-core) and three encoders (DINOv3 ViT-B/16, OpenPhenom, SubCell) under four community-standard protocols (replicate mAP, scIB batch integration, CellProfiler feature prediction, cross-batch perturbation recall), we find that the four protocols rank the same encoders systematically differently, with disagreements decomposing along three axes: cell versus background, morphology versus context, and within-study versus across-batch. The largest effect: on RxRx3-core, SubCell with segmented inputs retains 94% of crop replicate mAP but only 32% of crop R@10, so the within-study signal preserved under segmentation is largely non-transferable; density augmentation recovers 84% of the within-study C-to-S gap but only 8% of the cross-batch gap. Segmented views predict CellProfiler features as well as or better than crops on two of three datasets, inverting the replicate-mAP ranking, and the C-to-S gap varies by an order of magnitude across datasets while remaining similar across encoders, indicating that background-driven gain is set by experimental design rather than by the encoder. Single-metric ranking of Cell Painting encoders is therefore sensitive to the protocol used, and protocol disagreements are interpretable as projections onto the three axes the paired-view design exposes. We will release the paired-view datasets, reconstruction pipelines, 36 trained checkpoints, aggregated embeddings, and the full evaluation suite.
Read the full abstract

All authors Tim Treis, Nikita Moshkov, Johan Fredin Haslum, Shantanu Singh, Fabian J. Theis

Group Theory

A local approach to a programme of Meierfrankenfeld: initial setting and the symmetric case

A large $p$-subgroup of a group $G$ is a self-centralizing $p$-subgroup $Q \le G$ whose normalizer controls the normalizers of all the non-trivial, central subgroups of $Q$. In 2016 Meierfrankenfeld, Stellmacher and Stroth produced a result describing the $p$-local structure of a finite group having a large $p$-subgroup (the main examples arising from finite groups of Lie type in defining characteristic $p$). This result is a major success within the wider framework of studying groups of local characteristic $p$. We attempt to produce a result analogous to that of Meierfrankenfeld, Stellmacher and Stroth, but for fusion systems and localities. Previous work of Ellen Henke and the author shows that reasonable generalizations can be formulated and solved equivalently either within the realm of fusion systems or in the world of localities. In particular, in the present paper we set the stage for our analysis, showing how working within a locality grants a clear advantage: it allows to follow the same lines of reasoning as for a group. We therefore produce analogous reduction results and case subdivision as those in the 2016 result of Meierfrankenfeld et al. and, proceeding according to the analogy, we deal with occurrences of certain natural orthogonal modules and with the first of the cases that are to be studied, the so-called symmetric case. The remaining cases will appear in future publications.
Read the full abstract

Algebraic Geometry

The Chern character of a coherent sheaf on a smooth projective hypersurface

Given a coherent sheaf on a smooth projective hypersurface X, we prove an explicit formula for its Chern character as a Cech cocycle in terms of the free resolution of the associated module and calculate its image in the Jacobian ring under the Griffiths residue map. The formula is a geometric analogue of the Kapustin-Li formula for Landau-Ginzburg models, but proven directly using Hodge-theoretic techniques. This yields an effective method to compute the primitive part of the Chern character of any coherent sheaf using commutative algebra. We finish by proving the Hodge conjecture for the degree 33 Fermat fourfold.
Read the full abstract

Machine Learning

Curriculum-Based Adversarial Heterogeneous Agent Reinforcement Learning for Autonomous Quad-Copter Landing in Maritime Settings

Recovering unmanned aerial vehicles (UAVs) in maritime environments is challenging due to wind turbulence and ship-deck motion, making it a valuable test case for alternative control and learning approaches as conventional landing approaches often become unreliable. We study simulated mid-air capture of quadrotor UAVs by a ship-mounted robotic arm, learning robust cooperative control policies with Heterogeneous-Agent Proximal Policy Optimization (HAPPO) Reinforcement Learning. We train with HAPPO using a curriculum and an adversarial wind agent (HARL-AC) in NVIDIA Isaac Lab, and compare the obtained control policies against those generated through curriculum-based domain randomization and a benchmark trained on a single sea state. In-distribution evaluation on sea states $0/4/5$ shows comparable success for HARL-AC and domain randomization of up to $97.5\%$. On out-of-distribution sea states $7/8/10$, HARL-AC generalizes better, achieving up to $16\%$ higher median success rate at sea state 10, and substantially lower crash rates of up to $14\%$ compared to the domain randomization policy. Furthermore, we show that the adversarially trained policy shows more cautious behavior, slightly increasing timeouts by $<3\%$, but yields safer recovery behavior in severe, unseen conditions.
Read the full abstract

All authors Allan Minh-Tam Nguyen, Sree Showrya Kotala, Stefan Banioi-Crijman, Kurt Driessens, Rico Möckel

Software Engineering

GraphAHA: Graph-Based Adaptive Search with Heterogeneous Actions for Test-Time Code Generation

Test-time scaling improves code generation by spending additional inference budget (e.g., calls or tokens) on direct sampling, feedback-conditioned repair, and reasoning-guided implementation. Search-based methods can allocate this budget adaptively, but two challenges remain. First, tree-structured search treats each generation history as a separate state even when trajectories converge to the same program, duplicating evaluation and preventing statistics from being shared. Second, sampling, repair, and reasoning have complementary and state-dependent payoffs, making online allocation among them difficult under a finite budget. To address these challenges, we propose an adaptive graph search method with heterogeneous actions (GraphAHA). GraphAHA organizes the test-time code generation in a typed directed acyclic graph. Equivalent programs are merged into a single code node, allowing their downstream search statistics to be reused across all discovery paths. Hierarchical Thompson sampling then selects whether to generate a new state or follow an existing successor and, for generation, chooses among the type-valid sampling, reasoning, implementation, and repair operations. Evaluated on LiveCodeBench and CodeContests with Qwen2.5-Coder and DeepSeek-Coder, GraphAHA achieves the best score in 18 of 20 cases. For Pass@1 measured using visible tests, it outperforms the strongest baseline for both models on both benchmarks by 4.1 percentage points on average, demonstrating more effective use of a fixed inference budget.
Read the full abstract

All authors Xitao Li, Haijun Wang, Gege Yuan, Qiyuan Wu, Jiali Wei, Ming Fan, Xiaofei Xie