Analysis of PDEs
Papers
Fresh from arXiv, typeset properly. Highlighted words trade on the market.
All of arXiv
Tuesday 10 November
We study degenerate parabolic double phase systems with growth exponents $p<q$ whose modulating coefficient depends only on time. In this case, the operator on each time slice has Orlicz growth with structural constants independent of the coefficient, and the system is invariant under spatial translations. Using these features, we prove second order estimates and local bounds for the gradient when the coefficient is merely bounded and measurable, and $C^{1,\alpha}$ regularity when it is Hölder continuous in time. For weak solutions with gradients in $L^q$, these results hold without any gap condition between $p$ and $q$.
Robotics
GEMOR-GS: Global Extrinsic-calibration-refined Mapping and Online Relocalization for 3D Gaussian-Splatting SLAM
Photorealistic and metrically consistent world representations are a primary requirement for LiDAR–inertial–visual SLAM systems that serve robots over continuous operation. 3D Gaussian splatting makes such representations attainable by supplying an explicit, renderable, and queryable scene model. We present GEMOR-GS (Global Extrinsic-calibration-refined Mapping and Online Relocalization for 3D Gaussian-Splatting SLAM) to make the Gaussian map geometrically accurate, globally refined, and re-enterable, providing a better appearance for robot perception and decision-making. To improve the 3DGS map, the LiDAR–camera extrinsic is first estimated offline by direct-method alignment over dedicated calibration segments and reused by all following online and offline processes. Under the calibrated geometry, keyframe poses and Gaussian parameters are refined alternately under a gated differentiable-rendering and LiDAR–projection objective. The final global reconstruction is then rebuilt from the refined keyframe schema with all online constraints lifted. For the robot SLAM lifecycle, the method recovers the pose from the 3DGS map after tracking loss through an XFeat bypass map, RANSAC-PnP coarse localization, and Gaussian RGB-D warp refinement. Experiments on one self-collected and five public sequences quantify each element: the offline reconstruction wins all paired keyframes against the online map by 14.3–19.9 dB, improves PSNR by approximately 9 dB over Gaussian-LIC2, and achieves 86–100% held-out relocalization success tied to extrinsic quality. The results show that GEMOR-GS improves the geometric accuracy, appearance fidelity, and re-enterability of the 3DGS map, serving as a dependable robot world model for SLAM and downstream tasks.
Read the full abstract
Computers and Society
Repertoires of Contention: Building computational tools and infrastructures for community resistance to bitcoin mining
Recent work in human-computer interaction and allied fields has sought to engage community members as co-designers, facilitators, and decision makers to challenge structural power barriers and knowledge gaps. In this paper, we apply a diverse set of computational techniques to engage and address community concerns surrounding the socio-environmental impacts of the Greenidge Bitcoin mining power plant facility in upstate New York. Through collaborative interactions with community groups, addressing problems ranging from regulatory inertia to thermal water pollution, we aim to facilitate not just the production of new and critical environmental knowledge but also new computationally-supported repertoires of contention: means of contesting the unequal relations of knowledge and power that structure and often determine the outcomes of environmental controversies. We argue such engagements can help support more politically engaged forms of HCI, oriented towards efforts to redress and repair disparities in wealth, knowledge, and influence through strategic sociotechnical interventions and alliances.
Read the full abstract
Statistical Mechanics
A new representation of effective diffusion in fluctuating periodic potentials through stochastic duality
We revisit the classical problem of effective diffusion in fluctuating periodic force fields by combining homogenization with stochastic duality. We consider a specific class of fluctuations in which the internal state of the force field is independently redrawn from a fixed distribution at Poisson-distributed times. In the homogenization framework, the effective diffusion coefficient \(D_{\rm eff}\) is expressed in terms of a corrector function satisfying a backward Kolmogorov equation. For the redraw dynamics considered here, the equation governing the spatial derivative of the corrector can be reinterpreted as the stationary forward equation of a dual system obtained by reversing the force field, \(F\to-F\). This relation eliminates the corrector and yields a representation of \(D_{\rm eff}\) entirely in terms of stationary densities and probability currents of the physical and dual systems. We further introduce an auxiliary function \(Q(x)\) whose spatial variation is determined by the stationary current. When this current vanishes, \(Q(x)\) becomes independent of position and the effective diffusion coefficient reduces to the local relation \( D_{\rm eff}=L^2Q. \) This result generalizes the classical Lifson–Jackson relation to fluctuating periodic force fields. We verify the theory using analytical calculations, numerical solutions of the stationary equations, and direct simulations of the long-time particle displacement, including a driven system with nonzero stationary current.
Read the full abstract
Human-Computer Interaction
EJAI: Judgment-Preserving Civic AI for Data Center Environmental Justice
Communities in the United States and around the world are contesting data centers, the physical infrastructure of AI. These disputes pose environmental justice issues: residents must weigh unevenly distributed burdens using evidence scattered across metrics, corporate claims, policy documents, and public testimony. A formative study with 10 users showed that dashboards and single-agent AI can settle questions before people examine that evidence. Here we present EJAI, a source-grounded system that links an interactive map, retrieved evidence, missing-data statements, and two stakeholder personas that allow people to inspect evidence and form their own policy recommendations. In a small functional benchmark of 21 queries, retrieved evidence supported 98.4% of factual claims and the system acknowledged missing evidence in 32 of 33 cases. Persona alignment carried a grounding cost: the Corporate Executive omitted some evidence that conflicted with its assigned priorities. In a within-subjects study, 15 participants acting as residents prepared for a simulated town hall. Some checked cited AI claims while treating map metrics as official, some made conditional recommendations with safeguards such as water-use disclosure, and some found the personas simplified or biased. Our results suggest that civic AI can support preparation for public reasoning when it makes evidence, gaps, and disagreement open to inspection, and that corpus imbalance and two-sided personas remain open justice concerns.
Read the full abstract
Machine Learning
The RSI Tax: When Recursive Self-Improvement Is Not Worth the Cost
Recursive self-improvement (RSI) seeks to improve an agent using its own task experience. Its practical value depends on whether performance gains justify the full cost of self-improvement and task execution. We evaluate this trade-off on continual task streams, scoring each task before its feedback is used for subsequent updates. We define the RSI Tax as the token cost of the self-improvement process and include it alongside task-execution costs in end-to-end evaluation. Across six methods and two language models, a bounded-note baseline matches or exceeds the observed accuracy of Meta-Harness, EvoSkill, and RRSI, with mean total-token savings of 65.6-71.0%. We decompose the RSI Tax into verification, exploration, and writing costs, and track subsequent use of the associated state or artifact. For these three RSI methods, verification accounts for 74.8–93.6% of the aggregate RSI Tax. A substantial share of the RSI Tax is also associated with states or artifacts that are not used later in the evaluated streams. These results expose a practical limitation of costly RSI procedures: improvements in the deployed agent need not justify the expense of the improvement process. They motivate RSI objectives that jointly optimize the RSI Tax and expected task-execution cost while ensuring continued growth in agent capabilities across the task stream.
Read the full abstract
Machine Learning
Parallel Broyden Methods for Efficiently Evaluating Nonlinear State Space Models
Nonlinear state space models are typically evaluated sequentially, limiting their ability to exploit modern parallel hardware. Recent work has shown that state space models can be evaluated in parallel by reformulating evaluation as a root-finding problem, which can be solved with a parallel form of Newton's method. However, exact Newton methods require expensive multiplication of dense Jacobian matrices, and quasi-Newton methods based on diagonal approximations, while more scalable, are often slow to converge because they neglect interactions across state dimensions. We introduce a novel quasi-Newton approach based on Broyden's method, which captures coupling terms with diagonal-plus-low-rank approximations to blocks of the Jacobian. We construct the approximations from trajectory secants, avoiding automatic differentiation while retaining the favorable scaling of diagonal approximations and remaining compatible with parallel scan through rank compression. We prove finite-step recovery of the exact trajectory under the compressed scan, and show that convergence is tracked by the compression error and by the block approximation error along the current trajectory error, rather than in spectral norm. Empirically, we show that our method is advantageous when interactions are approximately low rank or when Jacobian evaluation is expensive: on low-rank RNNs, parallel MCMC algorithms, and denoising diffusion models, it converges in fewer iterations than other quasi-Newton methods and is the fastest parallel method when Jacobian evaluation is expensive or state dimensions are large.
Read the full abstract
Computational Physics
Fully GPU-accelerated semistochastic Epstein-Nesbet perturbation theory
Second-order Epstein-Nesbet perturbation theory (EN-PT2) is an important tool for improving selected configuration interaction (SCI) energies, but the large external determinant space creates substantial memory and communication costs on GPU clusters. We develop a fully GPU-accelerated implementation of distributed semistochastic EN-PT2, fusing excitation generation, screening, and PT2 weighting in a single kernel traversal. Fixed-size random-phase CountSketch arrays limit memory use and replace large external determinant exchange with bounded collective summation. The implementation on one A100 completes EN-PT2 calculations 8.21-37.53 times faster than Arrow on 64 CPU cores at the same requested statistical standard error, and fixed-workload strong-scaling tests achieve 93.5% parallel efficiency on 16 GPUs across two nodes. Applications to an iron-oxo complex in a (60e, 187o) active space, iron-porphyrin references containing 200 million determinants per state, and neural-network quantum states establish the method's efficiency for perturbative corrections to large reference wave functions in the GPU era.
Read the full abstract
Optimization and Control
A Fast Proximal-Gradient Algorithm for Decentralised Optimisation
We consider a connected multi-agent network where agents coordinate to minimise a finite sum of convex functions. Each function in the sum can be further expressed as the sum of a smooth and nonsmooth component and is known only to a single agent in the network. We propose a decentralised proximal-gradient-type algorithm for solving this problem, which we term Fast PG-EXTRA. In each iteration, agents evaluate the gradient of the smooth component of their local objective functions and the proximal operator of the nonsmooth component, followed by communicating the results with neighbouring agents in the network. Fast PG-EXTRA can be viewed as an accelerated variant of PG-EXTRA, in the sense that it enjoys non-ergodic $o(1/k^2)$ convergence rates in terms of the first-order optimality residual, successive iterates, and consensus error, compared to PG-EXTRA's ergodic $O(1/k)$ convergence rates. Our analysis of Fast PG-EXTRA is based on an accelerated variant of the Condat–Vũ algorithm for non-distributed convex minimisation, which has similarly improved convergence rates compared to the Condat–Vũ algorithm. We validate our convergence results for Fast PG-EXTRA and demonstrate its potential advantages compared to PG-EXTRA with numerical experiments for power state system estimation and $l_1$-regularised logistic regression.
Read the full abstract
Computers and Society
Developing an Operational Model for an Internet of Data Based Trusted Data Space for Vocational Education
Data circulation in vocational education has long been constrained by unclear ownership, a lack of trust, and insufficient incentives. The Internet of Data provides a new underlying technology for enabling trusted cross-domain data circulation. Building on an Internet of Data based trusted data space for vocational education, this paper develops a multi-stakeholder operational model encompassing a value network, a product portfolio, revenue allocation, and governance rules. Emphasizing data sovereignty protection and incentive compatibility, the model provides a reference for the sustainable operation of regional vocational education data ecosystems.
Read the full abstract
Astrophysics of Galaxies
Molecular signatures of shocked gas and enhanced cosmic-ray ionization in supernova remnant W49B revealed by ALMA
Galactic supernova remnants (SNRs) interacting with molecular clouds (MCs) provide important laboratories for investigating how supernova feedback is transferred to the interstellar medium. We performed observations of the HCO+ J=1-0, 3-2, and 4-3 lines, as well as the CO J=3-2 line, toward the eastern and western regions of SNR W49B using the Atacama Large Millimeter/submillimeter Array (ALMA) and the Atacama Pathfinder Experiment (APEX) telescope. In W49B west, we identify twelve regions with broad-line profiles in both HCO+ and CO emission, with FWHM of 25-60 km/s, providing clear evidence of SNR-MC interaction. These emissions are compact (at $\sim$5 arcsec resolution) and spatially coincident with 2.12~$\mu$m H$_2$ emission, but generally offset from the 1.64~$\mu$m [Fe~II] and radio continuum peaks. Non-LTE analyses yield shocked-gas densities of (0.5-1.6) $\times 10^5$~cm$^{-3}$ and high $N_{\rm HCO+}/N_{\rm CO}$ abundance ratios of (4.2-14.5) $\times 10^{-4}$. Reproducing such high abundance ratios in shocked MCs requires high cosmic ray ionization rates (CRIRs) of a few 10$^{-14}$-10$^{-13}$~s$^{-1}$, about 3-4 orders of magnitude above the standard Galactic level. Some regions exhibit enhanced HCO+/CO intensity ratios in their high-velocity line wings, possibly due to multiple unresolved shocks spanning a range of velocities and gas densities, spatial variations in the CRIR, or both. By contrast, the ratio remains nearly constant with velocity in other regions. For these regions, we further explore the propagation of low-energy CRs through the post-shock layer. An energy-budget analysis of a representative shocked clump indicates that the shock-associated kinetic-energy reservoir substantially exceeds cumulative CR and X-ray heating inputs, with CR heating exceeding X-ray heating.[abridged]
Read the full abstract
Astrophysics of Galaxies
The UV luminosity function at 1.6 < z < 2.2 from UVCANDELS
We present an analysis of the rest-frame UV luminosity function (LF) of galaxies at $1.6<z<2.2$ using the UVCANDELS F435W imaging data. Restricting our analysis to sources above $30\%$ completeness to minimize uncertainties at the faint end, we construct a catalog of 5,212 galaxies at $1.6<z<2.2$ down to an absolute magnitude of $M_{\rm UV} = -16.8$, using well-tested photometric redshifts. We perform a maximum likelihood estimate to derive the best-fit Schechter parameters of the UV LF at a median redshift of $z\sim1.9$. We report a best-fit faint-end slope of $\alpha = -1.586 \pm 0.042$. Splitting the sample into two redshift bins, we obtain $\alpha=-1.488 \pm 0.055$ at $z\sim1.75$ and $\alpha=-1.604 \pm 0.066$ at $z\sim2.05$ ($\sim1.4\sigma$ difference), alongside a $\sim2.3\sigma$ brightening in $M^{*}$. Integrating our best-fit LF down to $M_{\rm UV} = -15$, we derive $\rho_{\rm UV} = (1.821 \pm 0.037) \times 10^{26}\ \rm erg\ s^{-1}\ Hz^{-1}\ Mpc^{-3}$ at $z\sim1.9$. Combining our F435W measurements with previous UVCANDELS F275W results and other literature data, we model the redshift evolution of $\rho_{\rm UV}$ from $z\sim0$ to $z\sim10$.
Read the full abstract
Robotics
Differentiable Filtering for Inertial Odometry
Accurate inertial odometry is critical for autonomous systems but suffers from unbounded drift caused by the integration of sensor noise and biases. In approaches such as TLIO, neural pseudo-measurements are trained independently of their downstream filter. This paper introduces Differentiable Filtering for Inertial Odometry (DFIO), a learning framework that embeds the update stages of an error-state Kalman filter directly into a computational graph. Following standalone pre-training, DFIO jointly optimizes the displacement and covariance output heads using posterior-position supervision through differentiable measurement updates. The proposed system is validated on public pedestrian motion datasets featuring both calibrated and uncalibrated inertial measurements. Experimental results on a pedestrian dataset demonstrate that DFIO reduces mean absolute and one-second relative translation errors by 13.2% and 39.6% relative to the TLIO baseline, and by 7.7% and 17.4% relative to its pre-trained counterpart. Furthermore, evaluations on uncalibrated smartphone inertial data indicate that the proposed method effectively mitigates translational drift and maintains trajectory alignment under highly noisy sensor conditions. Source code will be made publicly available.
Read the full abstract
Networking
Confidence-Aware xApp Conflict Triage in Open RAN: A Closed-Loop ns3-oran Evaluation
In the Open Radio Access Network (Open RAN), independent RAN control applications (xApps) may request changes to the same radio parameter to optimize their own KPIs, causing conflicts. When conflicts are detected via key performance indicator (KPI) degradation, current conflict management systems usually resolve them immediately, without considering how long the degradation lasts or whether the network might recover naturally, wasting radio unit (RU) energy on interventions that may not have been needed. This paper presents a confidence-aware triage layer that assesses the reliability of its predictions by calibrating intervals from past prediction errors. It also decides, for each conflict, whether to tolerate, defer, or mitigate it based on the degraded KPI prediction confidence and persistence of the conflict. It is evaluated on a common conflict between two Open RAN xApps: coverage and capacity optimization (CCO) and energy saving (ES), which issue opposing requests for transmit power and affect reference signal received power (RSRP). The triage layer is validated in a closed-loop ns3-oran simulation with 5G-LENA. Results show that the method reduces unnecessary interventions by 30% and maintains RU energy efficiency relative to fixed priority while improving it by 2.6% over the no-mitigation baseline, while maintaining a statistical safety guarantee for RSRP.
Read the full abstract
Numerical Analysis
Multiscale LegONet: Learning Coefficient-Dependent Local Spaces for Composable Multiscale Finite Elements with Localized Error Bounds
We introduce multiscale LegONet (Ms-LegONet), a plug-and-play learned block for heterogeneous elliptic problems. One network per spatial dimension is trained once on local coefficient patches and reused, without retraining, across coefficient layouts, sources, boundary data and domain sizes within the trained coefficient class. On a fixed primitive patch the network maps the local coefficient to a zero-boundary space, the only learned object. The multiscale basis functions are the harmonic extensions of a fixed set of boundary traces, each a Galerkin solve in that space. Shared interface traces enforce continuity, larger blocks are assembled by recursive static condensation without any parent-scale network, and a localized Céa bound attributes the error of the learned spaces primitive by primitive. The training labels are harmonic responses of single patches, so no global solve enters the training. Against MsFEM, GMsFEM, the neural-operator element method and an end-to-end Fourier neural operator, Ms-LegONet is the most accurate at equal unknowns: its $L^2$ error is $4$ to $150$ times below the baselines on two-dimensional mosaics of $256$ high-contrast primitives, and its median error stays within $0.1$ percentage points of that of exact local spaces in three-dimensional layered media. For each new coefficient it is $1.6$ to $2.3$ times faster than a fine-grid direct solve in two dimensions. In three dimensions it computes the many-query port response of a chain $30$ to $900$ times faster than finite element solves on the same or a finer mesh, at a relative error of $10^{-4}$ against their $6\%$ and $1.6\%$.
Read the full abstract
Computer Vision
Seeing the Small: Deformable Duality Mamba for Retinal Lesion Segmentation
Diabetic retinopathy (DR) lesions vary widely in shape, and microaneurysms (MAs) span only a few pixels. We hypothesize that modeling the spatial organization of lesion regions, with global context and shape-adaptive scanning, improves the segmentation of such small and irregular lesions. To test this, we propose D$^2$Mamba, which couples non-causal state space duality with deformable Mamba scanning. We evaluate it with five-fold cross-validation on the DDR dataset, comparing against representative segmentation baselines under an identical training protocol, and conduct ablations on each component. D$^2$Mamba achieved the best overall performance, with 33.87% mIoU, 2.99 percentage points above the strongest baseline. The largest gain appeared on MAs, the smallest and most easily missed lesions, where Dice improved by 10.06 percentage points. These results confirm our hypothesis that spatially adaptive state space modeling benefits tiny-lesion segmentation. Code and pretrained weights are available at https://github.com/PUREPANEL1/D2Mamba.
Read the full abstract
Computer Vision
Evidential Deep Learning with conditional entropy for glaucoma detection
We propose a novel conditional entropy function as a basis for quantifying predictive uncertainty in Evidential Deep Learning (EDL) for glaucoma detection from retinal fundus images. Using the EyePACS-AIROGS-light-V2 and RIM-ONE-DL datasets for validation, we found that our proposed EDL model regularized by conditional entropy function produced uncertainty estimates with substantially lower calibration errors than Deep Ensembles, MC Dropout, post-hoc calibration and the conventional EDL framework based on Dempster uncertainty. To further probe our models' interpretability, we analyzed Grad-CAM heatmaps using a focus index representing spatial concentration. Compared with the other uncertainty quantification methods, EDL models exhibited greater separation between the index distributions of correctly and incorrectly classified fundus images. In particular, false negatives were associated with localized Grad-CAM activations that were displaced from the optic disc, consistent with its established clinical importance in glaucoma assessment. Overall, our findings suggest that our proposed EDL framework for glaucoma detection provides not only enhanced uncertainty calibration but also admits a more physiologically consistent interpretability.
Read the full abstract
Artificial Intelligence
Nepali-English Code-Mixed Speech Recognition: Data Preparation and Adaptation of a Speech Foundation Model
Nepali speakers regularly mix English words into conversation, but multilingual speech recognisers often struggle with language switches and the writing script used for English words. We study how to adapt SraVaani, a 449M-parameter hybrid TDT-CTC model, using a 172.8-hour corpus covering six Nepali speech domains. Its pretrained tokenizer does not support digits and several common symbols. Rather than replace it or discard these recordings, we generate Nepali and English readings for numbers and use zero-shot ASR output to select a plausible form. Full fine-tuning reduces test word error rate (WER) from 48.6% to 17.1%; a 12-epoch run reaches 16.3%, compared with 16.4% for fine-tuned Whisper large-v3-turbo. Freezing the encoder increases WER to 25.5%, while excluding numeric clips raises WER on numeric test speech from 14.3% to 21.3%. We also evaluate on FLEURS using both its original references and mixed-script alternatives. The comparison shows that transcription conventions explain part, but not all, of the out-of-domain performance loss.
Read the full abstract
Computer Vision
TMAL: Topology-preserving Multi-offset Affinity Learning for Curvilinear Structure Segmentation
Curvilinear structures such as retinal vessels and surface cracks need both pixel-level accuracy and structural continuity. Pixel-wise supervision does not model the connections between pixels, and topology-based losses still produce ordinary masks. Connectivity-based methods predict these connections directly, but they use binary connectivity targets over a fixed local range. Such labels give the same value to thin segments and to thick regions, and they cannot describe connections at larger offsets. To address these limitations, we propose Topology-preserving Multi-offset Affinity Learning (TMAL). TMAL uses soft affinities that emphasize thin regions near the boundary, and it predicts directional affinities at several offsets to reach further. It then combines these affinities with weights that come from the decoder features, so local continuity and wider context are used together. An auxiliary segmentation branch adds region-level supervision and keeps the overlap accuracy high. On the retinal vessel dataset DRIVE and the crack inspection dataset DeepCrack, TMAL reaches the highest Dice, IoU, and clDice scores, and it greatly reduces the connected-component error $\Delta\beta_0$. Code is available at https://github.com/suho-han/TMAL.
Read the full abstract
Computer Vision
Distribution Shift Detection in Remote Sensing Imagery using Machine Learning Models
Remote sensing imagery acquired by Unmanned Aerial Vehicles (UAVs) supports wide range of applications; however, deploying machine learning (ML) models on resource-constrained aerial platforms introduces challenges beyond predictive accuracy, particularly when deployment data differ from the distributions observed during training. Such shifts are common in remote sensing due to geographic, sensor, seasonal, and other visual variations. A practical solution is to equip deployed models with mechanisms that detect emerging distribution shifts and support subsequent responsive actions. This paper presents an empirical case study of three open-source detection tools: (i) Alibi-Detect; (ii) NannyML; and (iii) PyTorch-OOD. Using cross-dataset remote-sensing shifts, we evaluate both detection effectiveness and computational latency of these tools. Our results reveal differences in detection behavior and processing cost across methods and implementations, highlighting trade-offs between detection capability and response-time. These findings motivate further work toward lightweight, and latency-aware monitoring mechanisms for resource-constrained systems.
Read the full abstract
Computation and Language
Mobile AI Factory: Building Specialized Language Models for Your Mobile Device
Small language models (SLMs) running on mobile devices can keep user data private and work without connectivity. But they perform well only when fine-tuned for the target task and sized for the target device, a process that today requires repeated expert tuning. We present Mobile AI Factory, an agentic system that builds a fine-tuned, quantized SLM for a given task and phone. Given a task specification and a target phone, an orchestrator LLM chooses whether to tune hyperparameters, mine real data, or generate synthetic examples, while fixed rules roll back changes that hurt accuracy. The factory searches from small to large models across base models and quantizations, adapts its accuracy goal as it searches, and uses real measurements of the target phone across inference engines and backends. We evaluate Mobile AI Factory on ten tasks and four Android phones. On an unseen testset, its on-device models outperform two five-shot frontier models on five tasks, by up to 0.34, and match them on three more. Its search comes within 0.03 of exhaustively training all 27 model variants on an entity extraction task, while using 19$\times$ less GPU time. Phone measurements also matter: the best configuration changes across phones. Without phone measurements, Claude Opus 5.5, given the full hardware and model specifications, misses the app's requirements in 25% of easy and moderate cases and in 60% of hard ones. Code is available at: https://github.com/evanly-gh/SLM_Factory.
Read the full abstract
Mesoscale and Nanoscale Physics
Magnetization-free quantum anomalous Hall effect in altermagnetic weak topological insulators
The quantum anomalous Hall (QAH) effect is usually realized through ferromagnetism or magnetic dop ing, which limits material choices and often results in low mobility, disorder sensitivity, and low operating temperatures. Here, we propose a mechanism to achieve the QAH phase without net magnetization, using altermagnetic exchange on a weak topological insulator (WTI). The altermagnetic order selectively gaps the Dirac cone at the Z point, while hybridization between the top and bottom surfaces of the WTI thin film opens a trivial gap at the point, generating a Chern number C = 1 phase with a single chiral edge channel. We verify the topological transition through both a minimal thin-film model and a slab geometry, and compute the associated Hall response. As altermagnets and layered WTIs already exist experimentally, this mechanism offers a promising and experimentally accessible route to realizing QAH phases without magnetic doping or global magnetization. The results indicate that altermagnetic surface engineering may provide a clean and controllable pathway for realizing dissipationless edge transport in layered quantum materials.
Read the full abstract
Functional Analysis
Banach isomorphism between separable \(L_p\)-spaces, \(1\le p\ne2<\infty\), associated with hyperfinite type III von Neumann algebras
High Energy Astrophysical Phenomena
AGN feedback destroying group cool cores — a case study on IC 2476 with a remnant radio AGN and an exceptionally low abundance
We present results from deep XMM-Newton observations ($\sim$180 ksec) of the galaxy group IC 2476 in the X-GAP sample, capturing a rare AGN feedback regime, where the central radio AGN has ceased activity but the radio lobes are still detectable. The hot intragroup medium of IC 2476 features a compact cool core, with a size of ~10 kpc, and a moderate X-ray luminosity of $\sim3.4\times10^{40}$ erg/s. We measured an unusually low temperature of $\sim$0.45 keV within the cool core. We measured an exceptionally low abundance of below ~0.2 solar, even in the central region. The X-ray surface brightness of the group is not symmetric and shows a clear depression in the region overlaid with the eastern radio lobe, suggesting the presence of an X-ray cavity. The scaled entropy of the hot gas within 0.1 r500 of IC 2476 is among the highest in X-ray galaxy groups with a similar system temperature. Our study suggests that the past powerful radio AGN outburst in IC 2476 has disrupted the cool core and the X-ray cavity has dissipated or mixed with the surrounding hot gas since the shutdown of the central AGN. A large cool core may be gradually re-established in the absence of AGN activity. This system therefore provides a great example of AGN feedback destroying the group cool core, even with a moderate radio luminosity at present.
Read the full abstract
Earth and Planetary Astrophysics
A Possible Non-Transiting Companion in the TOI-2109 System Supported by Transit Timing Variations
Transit timing variations are a powerful method for detecting non-transiting planets, particularly in systems near mean-motion resonances. We investigate the ultra-short-period hot Jupiter TOI-2109 b using 87 transit epochs spanning two decades (2006-2025), combining observations from TESS, CHEOPS, HST, and ground-based photometry. We compare five timing models representing different physical scenarios: a linear ephemeris, orbital decay, apsidal precession, a sinusoidal model, and a two-planet n-body dynamical model. The orbital-decay fit yields $\dot{P} = -3.96 \pm 2.05$ ms yr$^{-1}$, providing no significant evidence for ongoing orbital decay. The sinusoidal model reveals a periodic transit-timing-variation signal with an amplitude of $\approx 44\,s$. Dynamical modelling favours an additional non-transiting companion near the 7:3 mean-motion resonance with an orbital period of $1.57813\pm 8.9\times10^{-5}$ days and a mass of $0.70 \pm 0.02\,M_{\rm Jup}$. The two-planet model yields $\Delta \mathrm{BIC}=4.95$ relative to the sinusoidal model, indicating positive evidence that the observed transit timing variations are best explained by gravitational perturbations from an additional planet.
Read the full abstract
Computer Vision
Dark Matters: VAE-Aware HDR Encoding for SDR-to-HDR Video Reconstruction
SDR-to-HDR reconstruction is an ill-posed problem in which clipped highlights, crushed shadows, and quantization artifacts must be recovered while preserving the original content. Latent diffusion models provide powerful generative priors for restoring missing detail, but their effectiveness is constrained by how HDR signals pass through the pretrained VAE. We find that common log-like encodings suppress low-luminance chromatic differences before generation, producing a dark-side collapse in the frozen-VAE latent. We introduce EquivFlow, a compact, invertible exposure-equivariant encoding that maps exposure changes to additive shifts while retaining a scale-independent chromatic component, without modifying the VAE. Encoding-level tests across pretrained image and video VAEs show that its exposure structure and deep-dark chroma transport persist across the tested hosts. Integrated into SDR-to-HDR video diffusion, EquivFlow delivers consistent chromatic-fidelity gains across the tested settings while retaining competitive reconstruction fidelity.
Read the full abstract
Sound
FaiSonic MMD: A popular music dataset with temporally aligned multidimensional annotations for multitask music analysis
Temporal analysis in music information retrieval concerns the positions of musical events and structures within recordings. Joint modeling has demonstrated the value of combining musical descriptions, but existing resources differ in joint coverage and expert verification across dimensions, constraining detailed joint supervision and the analysis of relationships on the same recordings. To support multitask modeling and the study of relationships among musical dimensions in popular music, we introduce FaiSonic Multidimensional Music Dataset (FaiSonic MMD), containing 500 songs totaling approximately 32.86 hours, comprising 250 Chinese-language and 250 English-language songs. Professional musicians annotate phrases, sections, chords, and rhythm in real recordings, followed by independent review. Events and intervals across dimensions share a recording-based time reference while retaining their respective musical granularity and boundaries, allowing direct comparison across musical descriptions. We describe the annotations, collection procedure, and statistics, and establish a comprehensive multitask baseline for phrase boundaries, section boundaries, section functions, chords, beats, and downbeats. We report reference performance for all prediction targets. FaiSonic MMD provides shared expert annotations and accompanying benchmarks for temporal music analysis.
Read the full abstract
Solar and Stellar Astrophysics
Thermal-pulse-resolved Binary Population Synthesis with Detailed Stellar Models: Implications for Long-period WD+MS Formation
The origin of the long-orbital-period white dwarf–main-sequence (WD+MS) binaries revealed by Gaia challenges the traditional association of common-envelope evolution (CEE) with short orbital periods. We develop a thermal-pulse-resolved binary population synthesis framework based on 51 detailed MESA evolutionary tracks with initial masses of 1–6 $M_{\odot}$. Applying these tracks to $5\times10^6$ initially sampled binaries, we follow the time-dependent stellar radius, wind mass loss, angular momentum, and tidal interactions, while calculating the envelope binding energy directly from the donor structure at the onset of Roche-lobe overflow. We find that thermal pulses can trigger binary interaction through envelope expansion and tidal orbital shrinkage, while the weakly bound envelopes of highly evolved thermally pulsing asymptotic giant branch (TP-AGB) donors allow CE ejection with limited orbital contraction. For our standard model, with $\alpha_{\mathrm{CE}}$ = 0.3 and $\alpha_{\mathrm{th}}$ = 0.5, systems with orbital periods above 30 days constitute approximately 11.5% of the resulting TP-AGB-initiated post-CE binaries (PCEBs). Long-period PCEBs can form across a large range of CE parameters and can reach periods of several hundred days even without any internal-energy contribution when $\alpha_{\mathrm{CE}}$ = 1. Their WD masses peak near 0.6 $M_\odot$, whereas systems containing more massive WDs are rare. Moreover, long-period PCEBs closely follow the single-star initial–final mass relation, providing a potential diagnostic of their formation channel in clusters, where progenitor masses can be independently constrained. Our results identify late TP-AGB CEE as a viable contributor to the observed long-period WD+MS population.
Read the full abstract
Classical Analysis and ODEs
Failure of endpoint estimates for lacunary bilinear spherical maximal functions on Heisenberg groups
We prove that the lacunary bilinear Nevo–Thangavelu spherical maximal operator on the Heisenberg group $\mathbb H^n$ does not map $H^1(\mathbb H^n)\times H^1(\mathbb H^n)$ into $L^{1/2,\infty}(\mathbb H^n)$ for any $n\geq1$. We obtain this result by lifting the Euclidean counterexample in our recent result (arXiv:2610.40100) to the Heisenberg group.
Read the full abstract
Mesoscale and Nanoscale Physics
Facilitated Relaxation Overrides Energetic Pathways in a Multiphase Square-Kagome Artificial Spin Ice
Relaxation in interacting many-body systems is typically assumed to follow energy minimization, yet local dynamical rules can determine which microscopic pathways are accessible. In facilitated relaxation, these pathways are enabled or blocked by neighboring configurations. However, isolating such rules in natural materials is difficult because energetics, disorder and dynamics are usually entangled. Here we realize a geometrically tunable square-kagome artificial spin ice-a lithographic nanomagnet array-that provides a clean, designable platform for facilitated relaxation. The array hosts two tunable nearest-neighbor interactions and supports multiple low-energy phases: spin crystal, spin liquid and spin foam. In the spin-foam phase, abundant facilitator configurations allow relaxation to follow Boltzmann statistics. In the spin-crystal phase, by contrast, domain-wall excitations are trapped in non-Boltzmann distributions, although energetically preferred relaxation channels exist. The ordered crystal background lacks the facilitators needed to connect topologically distinct domain-wall classes. These classes are therefore kinetically isolated, as observed in both experiment and Monte Carlo simulation. Our results show that transition-network topology can override energetic relaxation, establishing artificial spin ice as a programmable platform for constrained dynamics and nonergodic relaxation.