Peer Revenue Play credits. Real papers.

Papers

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

All of arXiv

Tuesday 8 September

Numerical Analysis

Third-order Halley-type iterative method with positive and bounded correction function

In this paper, we propose a new one-point iterative method for finding simple roots of nonlinear equations by modifying Halley's method. Its correction function is positive and bounded on $\mathbb{R}$ and matches the local expansion of Halley's correction function near a simple root. The method avoids the singularity and sign reversal of Halley's correction function while retaining at least third-order local convergence. We also establish sufficient conditions for global convergence. Numerical experiments show that the proposed method achieves high convergence success rates and favorable iteration counts over a wide range of initial guesses. Further experiments near the singularity of Halley's correction function demonstrate robust convergence behavior of the proposed method. An application to the van der Waals equation illustrates the effectiveness of the proposed method.
Read the full abstract

High Energy Astrophysical Phenomena

From DAMPE to LHAASO: Rigidity Scales and Composition Changes in Galactic Cosmic Rays

This commentary briefly reviews recent measurements from DAMPE and LHAASO on cosmic rays to explore why the spectra of different species change shape across the TeV-PeV energy range. DAMPE finds that several primary nuclei show spectral softening at a similar magnetic rigidity, while the combined proton and helium data from DAMPE and LHAASO reveal a more complex evolution at higher energies. By comparing these spectral features and composition changes, we discuss what they may tell us about the sources of Galactic cosmic rays and how these particles propagate through the Galaxy.
Read the full abstract

Computer Vision

SynthGait-19K: A Physically Grounded Synthetic Video Dataset for Gait Parameter Estimation

Accurate estimation of clinically meaningful gait parameters from monocular video is important for scalable mobility assessment, yet progress is limited by the small scale, restricted viewpoints, and limited visual diversity of existing datasets. We introduce SynthGait-19k, a physically grounded synthetic video dataset containing 19,272 walking videos derived from 6,427 MoCap sequences across 437 subjects, with paired SMPL motion and annotations for six gait parameters. To construct the dataset, we develop Gait2Vid, which unifies heterogeneous MoCap recordings through SMPL and synthesizes diverse RGB walking videos under controllable viewpoints and scene appearances. We assess the generated videos for consistency with their conditioning gait kinematics and validate extracted gait events against force-platform measurements. Using SynthGait-19K, we benchmark direct RGB, pose-based, biomechanical, and human-mesh-recovery approaches and analyze viewpoint, training-data scale, and synthetic-to-real domain shift. We also introduce GaitXFormer as a direct RGB reference model for estimating gait parameters. Synthetic supervision transfers effectively to real videos across both GaitXFormer and a pose-based architecture, demonstrating utility across different representations. We further find that spatial gait parameters are more sensitive to visual domain shift and that improved HMR reconstruction alone does not necessarily translate to improved downstream gait estimation.
Read the full abstract

All authors Soroush Mehraban, Xin Lei Lin, Vida Adeli, Majid Mirmehdi, Amirhossein Dadashzadeh, Clint Hansen, Andrea Iaboni, Babak Taati

Dynamical Systems

Dynamic Bifurcation of Nonautonomous Evolution Equations: Invariant Manifold Methods

In this paper, we use the global invariant manifold and the reduced singular cohomology groups method, which is different from those in the literature, to study the dynamic bifurcation from infinity of the nonautonomous evolution equation in terms of invariant sets by applying the Conley index theory (due to Rybakowski). We first establish a nonautonomous global invariant manifold for the abstract equation, which allows us to reduce the original system to this finite-dimensional manifold. Then, a homotopy between the reduced equation and a product flow is constructed. Finally, by considering the reduced singular cohomology theory of the Conley index, we establish our main theorems on dynamic bifurcations from infinity for this nonautonomous equation. As an example, a nonautonomous parabolic equation on unbounded domains is considered. Some new detailed results on bifurcations from infinity of the parabolic equation under an appropriate Landesman-Lazer type condition are proved, including the existence of a nonautonomous Morse decomposition, and improving the earlier works in the literature.
Read the full abstract

Machine Learning

Nyström Attention Matches Full Attention for Cross-Sectional Stock Prediction

MASTER's inter-stock multi-head attention – the module responsible for modeling cross-sectional stock relationships – accounts for 42.5% of model parameters and 25% of predictive value. We systematically decompose this module and uncover a surprising structure: the learned attention is near-uniform (perplexity 278/300), yet forcing exact uniformity eliminates all cross-sectional discrimination. Spectral analysis resolves this paradox: the deviation from uniformity is low-rank (effective rank ~65, top-10 modes capture 96.5% of energy), explaining why sparse approximations consistently fail while Nystrom low-rank attention (m=32 landmarks) matches full O(N^2) attention at O(mN) cost – certified equivalent via TOST at both N=300 (5 seeds, Rank IC p=0.003) and N=800 (10 seeds, Rank IC p=0.034). Additional findings include: (i) attention anti-correlates with return similarity (Spearman rho = -0.614; on the industry-labeled subset, -0.645 unconditionally and -0.627 after controlling for industry, beta, and volatility), suggesting complementarity-seeking rather than correlation mining; (ii) all graph-based alternatives degrade performance, with hard masking worse than complete module removal; and (iii) at N ~ 3,500 with adapted architectures, no cross-stock module (GCN, Nystrom, or MASTER-style pipeline) significantly outperforms a per-stock LSTM baseline (n=4 seeds), indicating that the benefits observed at smaller scales do not trivially transfer. These results establish that the inter-stock attention's value resides in a compressible, dynamic, near-global redistribution that rewards low-rank approximation but resists sparsification.
Read the full abstract

Artificial Intelligence

Artificial Intelligence-Assisted Digital Inventory of Cultural Heritage & Traditional Knowledge: Case for Indonesian Open Digital Library of Culture

The Indonesian Digital Library of Culture (Perpustakaan Digital Budaya Indonesia, PDBI; budaya-indonesia.org) is a participatory platform that has collected tens of thousands of entries on Nusantara cultural heritage through public contribution since 2007. Manual contribution faces three structural barriers: coverage (knowledge is scattered across languages and sites), integrity (open sources mix authentic documentation with noise), and completeness (subjects are recorded but their data remain shallow). This paper presents a methodological framework for autonomous, AI-based harvesting of cultural knowledge from the open web, designed to expand corpus coverage while intensifying per-entry data depth. The methodology is organised as a five-stage economic funnel: focused crawling, multilingual extraction and canonicalisation, vector encoding with blocking, agentic decision-making, and idempotent publication, under the principle of deterministic orchestration, agentic decisions. Each stage is formalised: funnel economics and optimal filter ordering; crawl-frontier dynamics as a subcritical branching process that explains the necessity of recurrent re-seeding; fact-level novelty via a containment measure; Bayesian multi-source evidence fusion with elevated publication thresholds for sacred categories; exactly-once effects via idempotent upserts and the transactional outbox; sliding-window inference budgeting with a reservation protocol; statistical quality auditing; and seed selection as submodular coverage maximisation. The framework retains four high-value human roles: curator of direction, escalation approver, quality auditor, and guardian of meaning, while machine autonomy is raised in stages. Ethical, legal, and cultural-sensitivity implications are discussed, including the architectural guarantee that the machine never overwrites human contributions.
Read the full abstract

Mesoscale and Nanoscale Physics

Tunable topological narrow bands in twisted bilayer-trilayer graphene

We investigate the low-energy band structure and topology of twisted bilayer–trilayer graphene with four stacking configurations: AB–ABC, BA–ABC, AB–ABA, and BA–ABA. Using both tight-binding and continuum models, we first establish that the two approaches show good agreement in the band structure in low-energy regime. We then study the evolution of the flat bands and their valley Chern numbers as functions of twist angle, perpendicular electric field, and the self-consistent Hartree potential. At relatively large twist angles and under electric field, we find a topological transition between the narrow bands, with the total Chern number of the flat bands following the Chern number sum rules derived from the chiral-limit description. We also observe another type of topological transition when the flat bands hybridize with adjacent remote bands, where gap closing and reopening processes lead to Chern number and charge density transfer. By constructing topological phase diagrams in the space of twist angle and electric field, we show that the perpendicular electric field provides an efficient tuning knob for controlling the stability and transitions of the Chern bands. Finally, we find that the Hartree potential mainly induce weak band shifts and reshaping in the narrow bands. However, with a combination of Hartree potential and the electric fields, the narrow bands show rich topological phase diagram. Our results clarify the interplay between the stacking, twist angle and electric field in manipulating the narrow bands and their topology in twisted bilayer–trilayer graphene, and provide guidance for engineering topological narrow bands with tunable Chern numbers in realistic twisted multilayer graphene systems.
Read the full abstract

All authors Dong Wang, Federico Escudero, Zhen Zhan, Shengjun Yuan

Cryptography and Security

AVP-Inspect: Coordinated Cyber-Physical Testing for Privacy Analysis of COTS Apple Vision Pro Applications

XR devices introduce substantial privacy concerns due to their comprehensive data collection capabilities that surpass traditional computing platforms. While existing works have demonstrated privacy concerns on Android-based XR devices such as Meta Quest series by performing network traffic analysis, little attention has been paid to the Apple Vision Pro (AVP) devices, mainly due to the closed nature and the technical challenges associated with AVP devices. In this work, we make a bold attempt to detect privacy violations of AVP applications from network traffic through automatic testing on AVP devices. Our key insight is that effective AVP application testing requires coordinated control of both cyber (software) and physical (hardware) components, which we term Coordinated Cyber-Physical Testing. Building on this insight, we design and implement AVP-Inspect, an automatic dynamic analysis framework for AVP applications, overcoming significant challenges enforced by the closed-source nature of AVP ecosystem. AVP-Inspect consists of three components: an automatic device controller by building customized hardware devices, a 3D UI explorer by designing a new exploration engine, and a privacy violation detector by constructing a unified privacy taxonomy for AVP. We first evaluated AVP-Inspect on a manually constructed ground truth dataset, then performed a large-scale analysis on 324 AVP applications downloaded from the App Store, with each app tested for 20 minutes. We found that 188 (58.0%) of apps exhibit at least one violation, and more than 60% of the network traffic flows are not properly disclosed.
Read the full abstract

All authors Yichang Xiong, Vamsi Shankar Simhadri, Yue Xiao, Xiaokuan Zhang

Machine Learning

Learning Metamaterial Eigenmodes with Wavelet-Encoded Fourier Neural Operators

Machine learning surrogates based on neural operators have shown broad applicability in solving forward PDE problems. However, eigenvalue problems, in which an eigenparameter and one of several valid eigenmodes must be simultaneously solved, remain difficult because standard operator learning formulations assume a unique input-output map. This work demonstrates that Fourier Neural Operators (FNOs), combined with wavelet-based encodings of PDE inputs, can learn and predict multiple eigenmodes of the elastic wave equation, corresponding to deformation modes of acoustic waves propagating through arbitrary metamaterial geometries. We provide a mechanistic explanation and experimental evidence for why wavelet encodings are well matched to the dual spatial-spectral structure of the FNO, enabling deterministic mode selection on both continuous-valued and binary-valued geometries within a single model, and for why prediction accuracy varies with geometric discontinuities. For metamaterial design, the resulting surrogate accelerates the simulation stage of the design cycle by three orders of magnitude relative to finite element analysis on a consumer-grade CPU, while preserving high fidelity. These results also carry broader implications for designing input encodings in other multi-mode PDE solvers based on spectral neural operators.
Read the full abstract

All authors Han Zhang, Alexander Ogren, Cynthia Rudin, Johann Guilleminot, L. Catherine Brinson

Biomolecules

PocketVE: Stable and Property-Guided Structure-Based Drug Design with Variance-Exploding Diffusion

Protein-conditioned 3D molecule generation is a central challenge in structure-based drug design, requiring a balance between pocket compatibility, molecular properties, and physical geometry. We propose PocketVE, a protein-pocket-conditioned variance-exploding (VE) diffusion framework that couples stable coordinate denoising with inference-time property guidance. Specifically, PocketVE combines an EDM-style training and sampling setup for 3D denoising, classifier-free guidance for multi-property steering, and adaptive protein perturbation as a training-time pocket regularizer.On the CrossDocked2020 benchmark under GenBench3D, PocketVE improves Valid$_{3\text{D}}$ from 58.6 to 80.6 and reduces strain energy from 457.4 to 127.9 relative to its guided TAGMol architectural parent; relative to TargetDiff, it attains comparable Valid$_{3\text{D}}$ with lower strain energy (127.9 vs.\ 306.0), while retaining competitive docking and molecular-property scores under moderate guidance. A guidance-scale study shows that moderate guidance gives a favorable balance between target-related objectives and geometric quality, whereas stronger guidance can degrade geometry and distributional fidelity. Pocket-permutation and PoseCheck diagnostics further support pocket-specific spatial compatibility with reduced steric conflicts. Overall, the results suggest that geometric stability and inference-time property guidance should be considered as coupled design objectives.
Read the full abstract

Mesoscale and Nanoscale Physics

Electronic Orbital Angular Momentum Driven by Finite-Momentum Phonons: Beyond Chiral and Axial Phonons

We show that finite-momentum phonons generate electronic orbital angular momentum (OAM) even without phonon axiality or chirality, with the response controlled by the phonon wave vector $q$ and frequency $\omega$. We develop a general gauge-field theory in which a unitary transformation absorbs the phonon displacement into emergent vector and scalar potentials acting on the electrons, providing a perturbative classification of the OAM response in $q$ and $\omega$. We derive $q$- and $\omega$-scaling laws for AC and DC responses. Notably, AC OAM arises even for linearly polarized phonons, with its magnitude and sign tunable by $q$. Its generation is governed by the matching between the phonon geometry and the electronic orbital texture rather than solely by the phonon angular momentum. The connection of the DC response to the electronic Berry curvature further supports that phonon angular momentum is not the only degree of freedom governing electronic OAM generation. Time-dependent tight-binding simulations under acoustic-phonon driving independently confirm the core predictions. Our results extend phonon-driven OAM beyond chiral and axial phonons and establish a wave-vector-tunable route to orbitronics, accessible with surface acoustic waves.
Read the full abstract

All authors Yongho Park, Jeonghun Sohn, Sejoong Kim, Kyoung-Whan Kim

Quantum Physics

Spectral Sign Implication for Quantum Logic

We study the spectral sign implication for quantum logic, defined by the nonnegative spectral projection of the operator obtained by subtracting the antecedent projection from the consequent projection. The construction agrees with classical material implication on commuting projections and compares arbitrary pairs of projections through the spectral structure of their difference. It satisfies Hardegree's four minimal implicative conditions, his law of contraposition, and a falsity condition. It differs from the standard polynomial implications in that its value need not belong to the ortholattice generated by its arguments. Within a uniform class of Borel constructions for pairs of projections, the operations satisfying entailment, contraposition, and the falsity condition correspond exactly to measurable choices of spectral branch. In the continuous subclass, these three conditions determine the spectral sign implication uniquely. In finite dimensions, the operation is also the largest among the acceptance projections of optimal projective tests for Helstrom discrimination between subspace states with priors proportional to rank. These results connect quantum implication with the operator theory of two projections and binary quantum discrimination.
Read the full abstract

Networking

6SEVEN: System for EValuating IPv6 ENumeration algorithms

The vast, sparsely populated, and often ephemeral IPv6 address space makes discovering active addresses challenging. In response, the community has developed over thirty different IPv6 Target Generation Algorithms (TGAs). TGAs learn addressing structure from seed (training) datasets, build a representative model, and generate candidate addresses. Unfortunately, the existing literature employs a wide variety of input seeds, data cleansing, and metrics of success that prevent ready comparison. Toward making TGA evaluation consistent, we present 6SEVEN, an extensible framework that hosts TGAs as plugins and links them to shared data cleaning, probing, dealiasing, and result tabulation components. We port as 6SEVEN plugins eight popular TGAs and demonstrate a controlled case study. Our results show that the performance of these TGAs varies substantially due to factors independent of the main algorithm—especially the seed set composition, dealiasing, and tabulation procedures. We observe tradeoffs between desirable features of TGA performance, including yield and exploration. We envision 6SEVEN as an enabling community resource to advance the science of current and future TGAs, and, by extension, IPv6 measurement.
Read the full abstract

All authors Chase Kanipe, Erik Rye, Dave Levin, Robert Beverly

Statistical Methodology

Nonparametric heterogeneous causal mediation with orthogonal machine learning

Causal mediation analysis decomposes the total effect of an intervention on an outcome into a direct pathway and an indirect pathway transmitted through a mediator, but standard methods typically summarize these pathways using population average effects. In many applications, however, the indirect effect may vary substantially across individual profiles. We propose an orthogonal statistical learning framework for estimating heterogeneous causal mediation effects conditional on individual characteristics. The method constructs a class of weighted Neyman orthogonal losses motivated by influence function representations of weighted population average effects. These losses directly target conditional mediation estimands whose minimizers are locally insensitive to nuisance estimation errors. We implement the resulting learners under a two-stage meta-learning framework with regularized linear sieves as second-stage smoothers, and introduce a combination of targeted learning and orthogonal learning designed to improve stability when mediator density ratios are unstable. We establish $L^2$ and uniform limit theory and develop pointwise and uniform confidence bands. Simulation studies show that the proposed orthogonal learners reduce the mean integrated squared error by more than $50\%$ compared with existing model-based methods and provide computationally efficient inference in nonlinear settings. The CARDIA, PSACR, and STAR analyses reveal heterogeneous mediated effects across cardiometabolic, psychological, and educational settings.
Read the full abstract

All authors Jiaqi Tong, Yi Zhao, Bhramar Mukherjee, Fan Li

Fluid Dynamics

Coupled thermoacoustic resolvent analysis of a model two-stream coaxial combustor

We derived a resolvent operator to analyze the coupled flame-acoustic effect in a model two-stream coaxial combustor. The theoretical analysis accommodates both the hydrodynamic effect of the active flame and chamber acoustic effect, as well as their coupling effect. Utilizing the coupled thermoacoustic resolvent, we are able to identify the optimal forcing of different mechanisms and their corresponding optimal responses. When this new modeling tool is applied to a two-stream coaxial model combustor, it reveals distinctly different forcing-response characteristics in the coupled system compared to those in a purely hydrodynamic linear system. Acoustic energy peaks, indicative of resonance, are observed. Additionally, it has been found that the acoustic forcing does not maintain a rank-one property. Furthermore, the modes most receptive to the flame-acoustic coupling effect can be identified using a novel sensitivity analysis approach. Interestingly, by employing the coupled thermoacoustic resolvent, we demonstrate that forcing can be designed to negate or counteract perturbations caused by the forcing of another attribute, potentially aiding in the development of effective thermoacoustic control strategies.
Read the full abstract

Cosmology and Nongalactic Astrophysics

Probing Projection Effects in Optically-Selected Clusters with Velocity Dispersions Outliers

The evolution of galaxy cluster abundance is a powerful probe of cosmology, but its efficacy depends on our ability to understand and control for systematics in the selection and mass estimation of clusters. We study redMaPPer-selected clusters in the Cardinal simulation, which simulates the DES Y3 observational conditions. Specifically, we investigate the line-of-sight velocity distributions of cluster member galaxies in 12,272 simulated clusters. We find that a significant fraction of clusters, 35%, have velocity dispersions that are high compared to their richness with an offset of $\sim 400$ km s$^{-1}$ between the outliers and the main population. A similar population of velocity dispersion outliers was found in the DES Y3 data. Nearly all of the outliers (98%) in a low-richness subsample have non-Gaussian velocity distributions, and they are more likely than the main population to have significant velocity substructure. These clusters also tend to have higher redshifts than clusters with typical velocity dispersions, and they show a larger scatter between halo mass and richness. Overall, our results point to a significant contamination of redMaPPer-selected samples from line of sight structure, particularly at high redshifts and low richness. Comparison to the Buzzard simulations shows similar results.
Read the full abstract

All authors K. Graham, T. E. Jeltema, C.-H. To, V. Wetzell, Kyle W. Davis, Conghao Zhou, H.-Y. Wu, E.S. Rykoff, J. DeRose, Michel Aguena, R.H. Wechsler

Artificial Intelligence

CIVI: A Framework for Diagnosing Search Agent Failures in Civic Information

Large Language Models are increasingly deployed in public-sector settings, where incorrect guidance can cause irreversible harm. We introduce CIVI, the first framework for diagnosing search agent failures in civic information. Its benchmark instantiation jointly spans cross-national, interjurisdictional government contexts (federal, state, and local) and functional categories from an internationally adopted United Nations standard. We evaluate ten frontier search agents and find that none matches an attentive human baseline. Alongside accuracy, CIVI measures search invocation rate, selective no-search accuracy, and how often agents cite authoritative government sources. To perform this diagnosis, we introduce ARISE, which decomposes agentic search failures into four mutually exclusive modes, isolated via source-injection ablation. ARISE attributes 72.1% of all observed failures to retrieval-bound causes rather than to gaps in the models' parametric knowledge.
Read the full abstract

All authors Dingying Liu, Yunshun Zhong, Wentao Zhang, Yiyuan Li

Cryptography and Security

LLM-Based Penetration Testing in the Presence of Honeypots

Large language model (LLM) agents are increasingly employed for offensive cybersecurity tasks such as automated vulnerability discovery, reconnaissance, and penetration testing. This new capability also threatens one of the defender's most valuable tools: deception. Traditional honeypots rely on realism and obscurity to lure human or script-driven attackers into revealing tactics, techniques, and procedures (TTPs), but LLM-driven attackers can reason about heterogeneous artifacts and use the honeypot suspicion to guide target-selection decisions. We present a systematic study of honeypot-aware budget allocation for LLM attack agents. We formalize the attacker's problem as a budgeted decision process: an agent interacts with potential targets, consuming LLM execution budget during reconnaissance and exploitation, and must decide whether to (continue exploitation) or (skip) when honeypot suspicion arises. Our findings show that with the proposed detector-guided policy, LLM agent attackers can effectively allocate budget to compromise hosts in a host pool, highlighting the importance of dynamically allocating budget in a controlled mixed-host testbed. While defenses are beyond our present scope, we discuss implications for future adversarially resilient and adaptive honeypot design.
Read the full abstract

All authors Xinhong Xie, Piyush Nagasubramaniam, Neeraj Karamchandani, Sencun Zhu

Logic

A Bounded Finite-One Degree Whose One-One Degrees Form Exactly a Dense Linear Order

We construct a set $U\leq_T\emptyset''$ whose bounded finite-one degree, ordered internally by one-one reducibility, consists exactly of a countable dense linear order without endpoints. More precisely, \[ \left( \{[B]_1:B\equiv_{\mathrm{bfo}}U\},\leq_1 \right) \cong (\mathbb Q,\leq). \] This exact realization contrasts with two previous results. In earlier work, a bounded finite-one degree was constructed that contains a copy of $(\mathbb Q,\leq)$, but the same degree also contains an infinite antichain of one-one degrees and, more generally, embedded copies of all countable partial orders; thus the dense chain does not exhaust the degree. In a different direction, $m$-rigidity yields an almost-sure and comeager obstruction: for a measure-$1$ and comeager class of sets, the corresponding bounded finite-one degree contains an infinite antichain of one-one degrees and hence is not linearly ordered. The present construction shows that, despite this typical negative behaviour, exact dense linear order can occur. In particular, it gives an affirmative answer to Open Question~3 of Richter, Stephan, and Zhang. The proof has two main parts. A block homogenization construction produces a noncylindrical set $U$, a weak dyadic tower $(U_e)_{e\in\mathbb N}$, computable reservoirs of both colours, and a base absorption property. An abstract absorption-to-exhaustivity theorem then shows that every member of the bounded finite-one degree of $U$ is one-one equivalent to some finite autojoin $mU_e$, thereby yielding the exhaustive classification above.
Read the full abstract

Group Theory

Braids on the Stranded Cellular Automata Model

The Stranded Cellular Automata (SCA) model is a grid of cells such that each cell can contain 0, 1, or 2 strands, together with two cellular automata that control when and how strands turn and cross. It was developed to study patterns occurring in fiber arts. We define a notion of what it means for a braid, in the sense of an element of a braid group, to be represented by an SCA pattern, and provide several algorithms to determine when a braid has an SCA representation with certain additional properties.
Read the full abstract

Artificial Intelligence

RevalExo: A Functional Daily-Activity Benchmark for Inertial and Visual Locomotion Mode Recognition in Older Adults and Clinical Cohorts

Assistive devices for people with mobility impairments, such as powered exoskeletons, rely on accurate locomotion mode recognition to adapt control strategies and provide appropriate assistance during daily activities. However, public benchmarks are typically collected from healthy adults, lack temporally precise labels necessary for detecting mode transitions, or focus on a limited set of tasks. To support development and evaluation under realistic clinical constraints and daily mobility demands, we introduce RevalExo, a functional daily-activity benchmark for inertial and visual locomotion mode recognition. RevalExo is built around a standardized, clinically and ecologically validated daily-activity protocol reflecting the cumulative everyday mobility demands in ageing and clinical populations. The benchmark includes 27 participants across three cohorts: older adults without mobility impairments, stroke survivors, and older adults with probable sarcopenia. The full cohort was recorded with lower-body IMUs, while synchronized egocentric video was collected for a clinically feasible subset of 13 participants. RevalExo provides 10.1 hours of frame-level annotations across 11 locomotion modes, including 5.1 hours of paired inertial–visual recordings. We benchmark three challenges: unimodal and multimodal locomotion mode recognition across multiple horizons, cross-population generalization from older adults without mobility impairments to clinical cohorts, and vision-guided knowledge transfer to IMU-only models. Results confirm consistent gains from fusing inertial and visual inputs but reveal a substantial gap between general recognition ($\sim$93% F1) and recognition during transitions ($\sim$68% F1), alongside persistent challenges in cross-population generalization and cross-modal transfer. We release RevalExo to stimulate further research on these open challenges.
Read the full abstract

All authors Diwas Lamsal, Juha Carlon, Reinhard Claeys, Maxim Yudayev, Louis Flynn, Tom Verstraten, David Beckwée, Eva Swinnen, Mihai Bâce, Bart Vanrumste, Benjamin Filtjens

Solar and Stellar Astrophysics

Multi-Height Spectropolarimetric Signatures of Magnetoacoustic Waves in Solar Plage

We analyze high-cadence spectropolarimetric observations of solar plage obtained with the Visible Spectro-Polarimeter (ViSP) on the Daniel K. Inouye Solar Telescope (DKIST) to investigate low-chromospheric waves sampled by the Na I D1 5896 Angstrom and Ca II 8542 Angstrom lines. Observations were taken at an oblique viewing angle, providing sensitivity to transverse motions. Magnetic oscillations within plage elements exhibit significant line-of-sight (LOS) RMS amplitudes of approximately 9 G on average in the Na I D1 line, whereas the corresponding Ca II oscillations are only marginally above the noise level. Adaptive profile tracking provided line-core intensities and Doppler velocities, the weak-field approximation provided LOS magnetic fields, and non-LTE inversions provided mass densities and formation heights. Cross-spectral analysis shows velocity-intensity phase distributions coupled at -90 degrees, consistent with compressive oscillations. The velocity phases in Na I-Ca II indicate upward propagation and the phase speeds are predominantly sub-Alfvenic, consistent with slow magnetoacoustic waves. Although the slit geometry prevents a definitive identification of the tube eigenmode, the differing magnetic phase relationships suggest a height-dependent sausage-like behavior in the coupling between the magnetic and compressive perturbations. Additionally, our inferred wave fluxes are insufficient to heat active-region plage.
Read the full abstract

All authors Taylor Sitterson, João M. da Silva Santos, Joshua Bentley, Tom Schad, Momchil Molnar, Bart De Pontieu, Juan Martinez-Sykora, Alexander G. M. Pietrow

Computation and Language

Vectorizer: Vectorizing NumPy Programs with Shape-Guided Rewrite

NumPy is a widely used Python library for numerical scientific computing, known for its declarative APIs and its optimized implementations. However, writing efficient NumPy programs, which often entails using vectorized array operations instead of explicit Python loops, may not be straightforward. This can be difficult for programmers who are accustomed to imperative array traversal, especially when vectorized API invocations require careful reasoning about shapes, broadcasting, and advanced indexing. This paper presents a rewrite-based approach for vectorizing Numpy programs with explicit loops over array data. Our approach vectorizes loops from the inside out, using array shapes and dataflow analysis to guide a source-to-source transformation that replaces loop bodies with vectorized statements. Following a set of rewrite rules that are correct by construction, our approach is consistently fast. We have implemented the approach as a tool called Vectorizer and evaluated it on 150 benchmarks collected from prior work and Stack Overflow. The evaluation shows that Vectorizer vectorizes 142 of the 150 benchmarks directly and 2 more after minor changes to the original benchmarks, with only 0.53 seconds on average to rewrite each one. The resulting programs are, on average, 74.83x faster than the original loop-based implementations.
Read the full abstract

Chemical Physics

Construction of downfolded Hamiltonians from projective transcorrelation

Projective transcorrelation recovers the short-range electron correlation by a similarity transformation with a geminal function $f(r_{12})$, at the cost of an effective Hamiltonian containing a 3-body term. We replace the term by an effective operator of rank at most two, obtained from the two-body cumulant (2C) approximation to the three-particle reduced density matrix. The 2C 3-body energy is written without cumulants, and the effective one- and two-body interactions are derived as its partial derivatives with respect to the reduced density matrices. The truncation is assessed on the atomization and reaction energies of the HEAT set with CCSD(T), and on the CAS-pTC model, whose downfolded Hamiltonian is held on a qubit register with the number of Pauli strings reduced from the sixth power of the orbital count to the fourth.
Read the full abstract

Instrumentation and Methods for Astrophysics

CTree: Branch-wise compatible merger tree builder

We introduce a merger tree builder, CTree. Its key feature is the identification of progenitors using the evolutionary information encoded in object branches, rather than relying solely on the instantaneous particle membership in a single snapshot. In this sense, CTree can be described as a branch-wise merger tree builder. Specifically, CTree extracts particles associated with objects over a defined timescale and uses this information to compute the merit functions employed in the progenitor search across multiple snapshots. During this process, CTree requires only the identifiers of particles and objects. This ensures high compatibility with different halo and galaxy catalogues and also with pre-existing merger trees. To identify `core' particles that remain bound to the host halo over extended physical timescales, CTree employs a simple and efficient algorithm that does not rely on explicit radial or binding-energy information. The timescale adopted for core particle extraction and the number of snapshots used in the progenitor search can be user-configured. Basic validation tests using suites of dark matter-only simulations demonstrate that CTree produces stable and physically consistent halo merger trees, particularly in high-density regions or in simulations with high output cadence. The code is written in C++ and employs a hybrid message passing interface and OpenMP parallelization, making it well-suited for high-resolution cosmological simulations. The code is publicly available and actively maintained to improve performance, reliability, and compatibility.
Read the full abstract

Analysis of PDEs

Construction of three-solitons with logarithmic distance for the mass-critical gKdV equations

For the mass-critical generalized Korteweg-de Vries equation, \begin{equation*} \partial_{t}u+\partial_{x}\left( \partial_{x}^{2}u+u^{5}\right)=0,\quad (t,x)\in [0,\infty)\times \mathbb{R}, \end{equation*} we prove the existence of three-soliton solutions with logarithmic relative distance and with the choice of signs $(+,-,-)$. The choice of the number three and the signs of solitons are related to the solvability of the ODE system generated by the nonlinear interactions between the three solitons and some non-localized profiles. In particular, these special behaviors are due to strong interactions between the three solitons. That is, the dynamics of each soliton is perturbed at leading order by the presence of other solitons.
Read the full abstract

Computer Vision

Marigold V2: Revisiting Diffusion Transformers for Monocular Depth Estimation

Monocular depth estimation is a ubiquitous yet highly ill-posed computer vision task, with downstream applications in scene reconstruction, computational photography, and robotics, among others. Despite the field's maturity, recent models still struggle to generalize to out-of-distribution inputs and to produce sharp and detailed depth maps. In this paper, we revisit Marigold, a set of techniques for repurposing modern image generation and editing models, powered by the diffusion transformer (DiT) architecture, into state-of-the-art monocular depth estimators. Our recipes target single-step inference from pretrained multi-step flow-matching models, with quantization where needed, preserving model capacity while remaining cheap to run. We analyze the artifacts of naive training and identify two effective remedies: aligning the model's internal representations with semantic features extracted from ground-truth, and adopting a 2-stage fine-tuning protocol built around a novel Sinkhorn-based loss. The results are crisper, cleaner depth maps that generalize well out-of-distribution, with 16-26% improvement in AbsRel over the previous best on KITTI and ETH3D. Qualitatively, our model resolves fur, foliage, and hair-thin edges that have eluded prior models. Furthermore, Marigold V2 achieves state-of-the-art results when applied to other dense regression tasks, such as surface normals estimation and intrinsic image decomposition. Project website: https://hf.co/spaces/huawei-bayerlab/marigold-v2-web
Read the full abstract

All authors Igor Pavlovic, Thiemo Wandel, Anton Obukhov, Luca Bartolomei, Andrey Davydov, Fabio Tosi, Matteo Poggi, Sabine Süsstrunk, Dengxin Dai

Atomic Physics

A Portable Dual-Color Two-Photon Rubidium Optical Frequency Standard

Portable atomic clocks are essential in a wide variety of applications, most notably in the operation of global navigation satellite systems. Existing portable atomic clocks utilizing microwave-based interrogation schemes are now routinely eclipsed by the next generation of atomic frequency standards based on optical interrogation. While optical frequency standards demonstrate greatly improved frequency stability, they have only recently reached a level of technical maturity required to demonstrate this improved performance outside of well curated laboratory environments. Here, we demonstrate a fully autonomous and portable optical frequency standard based on an efficient dual-color excitation of the $5S_{1/2}\rightarrow5D_{5/2}$ two-photon transition in $^{87}$Rb. The standard utilizes a combination of robust, highly developed commercial-off-the-shelf telecommunications technologies and a fully integrated portable optical frequency comb, providing the optical and microwave outputs vital for interfacing with existing electronic systems and infrastructure. The system demonstrates a fractional frequency stability of $1.9\times10^{-13}$ at 1s of integration time, reaching $3.5\times10^{-15}$ at 8000s of integration time without the need for drift removal. This portable demonstrator unit marks a significant achievement in the development of Rb optical atomic frequency standards, and for the deployment of optical atomic frequency standards outside of the laboratory.
Read the full abstract

All authors Sarah K. Scholten, Emily Ahern, Clayton Locke, Christopher J. Billington, Nicolas Bourbeau Hébert, Montana Nelligan, Ashby P. Hilton, Rachel F. Offer, Elizaveta Klantsataya, Christopher Perrella, Andre N. Luiten

Artificial Intelligence

Inference-Time Nash Alignment

Preference-based fine-tuning methods such as RLHF and DPO require substantial compute and large preference datasets. They also need direct access to the model parameters which are not provided by many state-of-the art models. Inference-time alignment offers a cost-effective alternative without updating model parameters. However, existing inference-time methods rely on a scalar reward model derived under a Bradley-Terry assumption, which cannot represent general preferences. Following recent work on fine-tuning with generalized preferences, in this work, we initiate the study of inference-time alignment under general preferences. We formulate the problem as obtaining a Nash equilibrium of a two-player zero-sum game between policies. We propose two algorithms: Best-of-Nash (BoN) and Nash Mirror Descent (NMD). We prove that both algorithms achieve a duality gap that matches the problem lower bound. Empirically, we implement the two methods on three datasets, which shows that our methods substantially outperform the base policy, converging to the performance of the fine-tuned models. Moreover, our results show that NMD remains robust across the regularization parameter.
Read the full abstract

Image and Video Processing

Reliability assessment and multicenter clinical application of magnetic resonance methods for knee cartilage quantification

Background: This study evaluated interreader agreement and longitudinal performance of MRI methods for knee cartilage volume, thickness, and defect-area quantification. Methods: AI-presegmented masks from 1,189 phase III examinations underwent independent correction by two readers and adjudication. Cartilage volume, three-dimensional ray-tracing thickness (3D-RT), and ray-based defect area (3D-RBA), defined by a 1.5-mm thickness threshold, were calculated. Agreement was assessed using segmentation metrics, intraclass correlation coefficients (ICCs), repeated-measures Bland-Altman analysis, and minimal detectable change at 95% confidence (MDC95). The 3D-RBA framework was evaluated in 120 digital-phantom experiments from 40 participants. Longitudinal analyses included 374 participants, alternative-method comparisons included 65, and retrospective phase II analysis included 24 participants with four visits. Results: Overall AI-to-adjudicated-mask Dice was 0.964 +/- 0.029. Interreader ICCs for volume, thickness, and defect area were 0.956, 0.904, and 0.932; corresponding MDC95 values were 1,596.9 mm^3, 0.227 mm, and 147.4 mm^2. Geometric mean absolute percentage error for defect area was 5.62%, with spatial Dice of 0.961. In 374 participants, volume changes correlated positively with thickness changes (rho=0.431) and negatively with defect-area changes (rho=-0.221). Within-participant phase II correlations followed the same directions in both groups. Conclusions: The workflow demonstrated good interreader agreement. Controlled geometric results and longitudinal associations supported the feasibility of threshold-based defect-area estimation. Volume, thickness, and defect area provide complementary measures of cartilage structure.
Read the full abstract

All authors Binbin Yang, Yongmei Jian, Chenglei Liu, Rui Huang, Hongda Shao, Linjun Tong, Yuanjing Xu, Jingshu Wu, Chengzhang He, Suting Peng, Ming Xiao, Yinan Chen, Qi Duan