arXiv subjects

Ying Zhou

Publications and source records attributed to Ying Zhou.

At least 19 recordsLinked to original sources

arXiv2026

GRAFT-EEG: Gauge-Aware Physical Field Transport for Cross-Dataset Motor-Imagery EEG

Cross-dataset EEG decoding often treats montage, reference, and active-sensor changes as representation shift, although they also alter the observation operator. We study target-statistics-blind transfer: fitting receives no target signal, label, or empirical statistic, and electrode coordinates only instantiate a fixed prediction operator. method removes the active-set common-reference gauge and transports each native mask through a template motor leadfield. A source-fitted spectral classifier anchors the decision; a bounded state-space-model (SSM) residual is secondary. We establish restricted identifiability and an operator-perturbation bound. Across five leave-one-dataset-out targets (253 participants), dynamic method achieved 60.865% equal-target balanced accuracy versus 52.684% for strict-clean 5/5 BIOT (hierarchical difference +8.182 points; 95% CI +6.382 to +10.196; positive in every target). On four EEGPT-clean targets, the descriptive means were 61.520% and 59.057%. In four post-hoc exact same-parent tests, FULL exceeded coordinate-basis, matched-spectrum, row-shuffled, and wrong-anatomy controls by 9.959, 9.698, 11.241, and 3.985 points, respectively (all 95% CIs excluded zero; Holm-adjusted P<0.001). A separate 60-participant same-paradigm evaluation retained a positive physical-operator contrast (+4.029 points; +2.338 to +5.773), whereas dynamic--static was inconclusive; its preregistered status is partial confirmation rather than unrestricted external validation. The SSM added +1.059 points (+0.013 to +1.892). Claims remain bounded to five binary motor-imagery domains and the tested operators.

cs.LG↗

arXiv2026

TriRouter: Sparse-to-Dense Electrode-Space Routing for Robust EEG Motor Imagery Decoding

Motor imagery brain-computer interfaces require reliable decoding from multichannel EEG, but practical acquisition often yields sparse or corrupted electrode montages. We propose TriRouter, an electrode-space routing framework with discriminative-preservation, mask-robustness, and latent-compensation paths. A validation-calibrated router uses observable channel availability to allocate path contributions, preserving complete-channel discrimination while increasing representation recovery under severe missingness. In BNCI2014-001, TriRouter outperforms standard public backbones and a dynamic-spatial-filtering corrupted-EEG baseline. It improves accuracy over the strongest corrupted-EEG baseline by 3.68, 4.02, and 6.46 percentage points at nominal 25%, 50%, and 75% missing settings, and remains effective under structured regional missingness and additional low-channel dataset evaluation.

cs.LG↗

arXiv2026

EHR-Bench: A Chinese Multi-Task Benchmark for Recruitment AI

Recruitment applications span diverse operations, from drafting job descriptions and interpreting queries to organizing profiles, assessing fit, and recommending candidates and jobs. Systematic LLM evaluation therefore requires task coverage that reflects this range of business needs. We introduce EHR-Bench, an open Chinese benchmark comprising twelve tasks across six capability families and five recruitment scenarios. The tasks cover generation, extraction, classification, retrieval, semantic similarity, and reciprocal ranking within one evaluation framework. An internal recruitment dataset guides task definitions, the occupation taxonomy and sampling targets, output requirements, and scoring protocols. Following these specifications, we construct 8,336 task instances from public materials and controlled augmentation, with source-aware expert review and documented provenance. We evaluate twelve LLMs through task-level, capability-level, and aggregate comparisons. Closely ranked models exhibit distinct task strengths, and performance varies across input sources, reference evidence, and candidate pools. Five shared models retain the same observed Overall ordering on a private operational reference, with variation in capability and task correspondence. EHR-Bench provides a reproducible foundation for systematic model assessment, task-specific model selection, and further research on LLMs for recruitment. The project repository is available at https://github.com/EhrBench/EHR-Bench.

cs.CL↗

arXiv2026

MoSE: Mode-Switching Expander for Mixed LLM Training and Inference

AI clusters increasingly run large language model (LLM) inference and training on the same fabric. Prefill-decode (P-D) disaggregation creates key-value (KV) cache transfers between prefill and decode groups, whereas training collectives and all-to-all traffic benefit from near-uniform global connectivity. A static sparse topology can therefore be poorly matched to one of the two traffic patterns. We present Mode-Switching Expander (MoSE), a reconfigurable expander that treats topology design as a fixed-degree edge-allocation problem. MoSE reallocates the same sparse edge budget toward direct P-D connectivity in inference-heavy modes and restores a uniform random regular expander in training-heavy modes. We evaluate MoSE using a 1024-group flow-level topology model, shortest-path routing, and two mixed workloads. Across 20 seeds, MoSE reduces average and 95th-percentile (P95) load-aware KV communication cost by 90.8\% and 91.9\% relative to Static-Training in the inference-heavy mode. In the training-heavy mode, it reduces average and P95 training communication cost by 22.7\% and 27.6\% relative to stale Static-Inference. These results show that coarse-grained topology switching can support both workload modes without additional ports or routing changes.

cs.NI↗

arXiv2026

Optimizing VLP-aligned Multimodal Intent Representation with Correct Visual Instantiation for Zero-Shot Composed Image Retrieval

ZS-CIR aims to retrieve a target image from a reference image and a modification text without paired supervision, typically by encoding composed queries as text-dominant representations within the image-text matching space of VLPs. However, queries reconstructed by visual pseudo-word learning or MLLM-based target reasoning often deviate from the native VLP representation space due to reference noise and coarse text fusion in the former, and verbose, weakly visually grounded descriptions in the latter. In this paper, we propose a unified ZS-CIR framework (named VMIR-CVI) to reconstruct multimodal composite queries from two complementary perspectives for optimizing VLP-compatible multimodal intent representation. First, it reasons and converts the multimodal intent into a unified textual description, aligning with the native text space of the VLP backbones to produce more retrieval-compatible textual queries. Second, it reconstructs the query representation with correctly decoupled visual instance cues, reducing reference noise while preserving target-relevant content. Specifically, a VLP-aligned Multimodal Intent Reasoning (VMIR) module injects few-shot VLP-style exemplars into chain-of-thought prompts, guiding the MLLM to generate target-consistent intent queries. A Training-free Visual Instance Disentanglement (TVID) module decouples fine-grained visual instances from global reference features without additional optimization. Finally, a lightweight Hybrid-modal Intent Alignment and Fusion (HIAF) module integrates the reasoned textual intent and disentangled visual cues into a unified hybrid-modal representation for robust ZS-CIR. Extensive experiments on three CIR benchmarks, namely CIRR, CIRCO and FashionIQ, show that VMIR-CVI significantly outperforms existing baselines and achieves new state-of-the-art performance. Code and trained models will be publicly released.

cs.IR↗

arXiv2026

WebCraftBench: Evaluating Web Application Generation from a Software Testing Perspective

Human evaluation provides a direct measure of the quality of LLM-generated web applications. However, fitting human judgments through automated evaluation remains challenging. Static benchmarks can credit functionality that exists in source code but is unreachable at runtime. Interactive benchmarks exercise the application, yet incomplete exploration can cause them to miss implemented functionality and confound application defects with agent execution failures. To address these limitations, we propose WebCraftBench, an interactive benchmark for evaluating web application generation from a software testing perspective. WebCraftBench instruments each generated application and uses code coverage to guide an agent in exploring its functionality through user-simulated interactions. It then abstracts the interaction trace into a state-transition graph and evaluates the application along three dimensions: visual aesthetics, usability, and requirement alignment. By separating exploration from scoring, WebCraftBench collects runtime evidence without constraining exploration to predefined acceptance criteria. WebCraftBench comprises 369 real-world user requirements and 5,088 acceptance criteria. Evaluation of 17 frontier LLMs reveals distinct strengths across the three dimensions, with no model leading on every dimension. On 197 validated sessions sampled from an internal arena, WebCraftBench achieves 85.3\% agreement with human preferences, with agreement generally increasing as the score difference between paired applications grows. Further experiments show that coverage guidance improves exploration coverage and the model rankings remain stable when the judge model is replaced.

cs.SE↗

arXiv2026

Soft Posterior Speaker Injection for Multi-Talker Speech Recognition

Multi-talker automatic speech recognition (MT-ASR) remains challenging in the presence of overlapping speech. Hard segmentation introduces irreversible errors, whereas serialized output training (SOT) avoids explicit segmentation but does not condition a pretrained encoder on speaker activity. We propose Soft Posterior Speaker Injection (SPSI). A Soft Posterior Head predicts per-frame speaker posteriors $\hat{\mathbf{P}}$ and injects them into Whisper through Multi-layer Feature-wise Linear Modulation (MFLM) and Speaker Memory Prompts (SMP). The benefit of SPSI is largest where overlap is heaviest and under domain transfer. On controlled two-speaker LibriSpeech overlap, SPSI reduces concatenated minimum-permutation word error rate (cpWER) from $61.5\%$ to $60.0\%$ in the high-overlap bin, and from $51.9\%$ to $51.0\%$ on the full set, relative to SOT. By contrast, Speaker CE, SD-CTC, SA-DiCoW, and Pipeline (oracle/est.\ VAD) do not outperform SOT. Freeze-posterior overlap-heavy adaptation reduces held-out LibriCSS cpWER from $42.3\%$ to $36.8\%$ on sessions $8$--$9$, a $5.5$-point gain over SOT. The source code is available at https://github.com/HackerHyper/SPSI.

cs.SD↗

arXiv2026

Identifying and Transferring Reasoning-Critical Neurons: Improving LLM Inference Reliability via Activation Steering

Despite the strong reasoning capabilities of recent large language models (LLMs), achieving reliable performance on challenging tasks often requires post-training or computationally expensive sampling strategies, limiting their practical efficiency. In this work, we first show that a small subset of neurons in LLMs exhibits strong predictive correlations with reasoning correctness. Based on this observation, we propose AdaRAS (Adaptive Reasoning Activation Steering), a lightweight test-time framework that improves reasoning reliability by selectively intervening on neuron activations. AdaRAS identifies Reasoning-Critical Neurons (RCNs) via a polarity-aware mean-difference criterion and adaptively steers their activations during inference, enhancing incorrect reasoning traces while avoiding degradation on already-correct cases. Experiments on 10 mathematics and coding benchmarks demonstrate consistent improvements, including over 13% gains on AIME-24 and AIME-25. Moreover, AdaRAS exhibits strong transferability across datasets and scalability to stronger models, outperforming post-training methods without additional training or sampling cost.

cs.CL↗

arXiv2026

Tracing Query Expansion Effects through Sparse Autoencoder Features

Query expansion (QE) is a critical technique in information retrieval that enriches underspecified queries with additional textual context. However, its effect is often unreliable in modern dense retrieval, especially for strong off-the-shelf retrievers without retraining. Existing studies mainly examine expansion quality, semantic drift, or retrieval outcomes, but rarely explain how QE changes dense retrievers internally. In this work, we trace QE effects through sparse autoencoder (SAE) features. Using paired original and expanded queries, we decompose layer-wise retriever representations into sparse latent activations, identify QE-related latents from expansion-induced activation shifts, and interpret them with natural-language descriptions and retrieval cases. Our analysis shows that effective QE induces layer-concentrated changes in sparse latents aligned with retrieval intent and entity attributes, rather than only perturbing final query embeddings. SAE-based activation steering further validates these latents improve retrieval more consistently than random interventions or vanilla QE across four benchmarks, suggesting that SAEs can explain QE effects and offer a lightweight option for precise retrieval behavior modulation without query rewriting or retriever fine-tuning.

cs.IR↗

arXiv2026

Locally Private Subgraph Counting via Noisy Adjacency Matrix and Differential Privacy on Randomized Data

Subgraph counting is a fundamental primitive for graph analytics, with applications ranging from social recommendation to anomaly detection. Private subgraph counting under edge local differential privacy (edge-LDP) is challenging without a trusted server or shuffler. We propose two abstractions: the \textbf{Noisy Adjacency Matrix (NAM)} framework, which reformulates private subgraph counting as algebraic estimation over unbiased randomized matrices, and \textbf{Differential Privacy on Randomized Data (DPRD)}, which gives a composition rule for calibrating later-round noise when auxiliary inputs are randomized. Using these tools, we design one-round and two-round estimators for triangle and quadrangle (4-cycle) counting, including TriOR, TriTR, TriMTR, QuaTR, and $\text{TriTR}^*$. We provide the first exact closed-form MSE for the RR-based one-round triangle estimator class and the first relative-error bounds for two-round triangle counting, showing error independent of graph size when the clustering coefficient and average degree are stable. Experiments on two real graphs show that the proposed two-round algorithms outperform all evaluated edge-LDP and shuffle-model baselines in terms of relative error.

cs.CR↗

arXiv2026

Numerical modeling and simulation on nanosecond laser-target interactions

Nanosecond lasers are widely used in industrial applications as they are relatively inexpensive, and their compactness and robustness are an advantage. Much experimental work has been carried out to understand deeper the interaction between the nanosecond laser pulses and the targets, as these are complex, transient processes with spatial inhomogeneities. Beside the experiments, the modeling and numerical simulation on the laser interaction with the target are also crucial for understanding the dynamics of laser-material interactions and for optimizing laser processing applications. In this review, the progress of numerical modeling and simulation on nanosecond laser-target interactions are summarized from the aspects of laser-target interactions and target-plasma interface, laser-plasma interactions and plasma radiation, and numerical models on different scales with artificial intelligence advancing. The laser ablation, mass and energy transfer, and mechanical coupling are discussed in the aspect of the nanosecond laser-target interactions and target-plasma interface. The plasma expansion, plasma ionization and recombination, and plasma radiation are discussed in the aspect of the nanosecond laser-plasma interactions and plasma radiation. Then the numerical advances, including microscopic approaches based on molecular dynamics, mesoscopic approaches based on kinetic and statistical physics, macroscopic approaches based on fluid dynamics, and numerical simulations with machine learning are discussed. Finally, the challenges currently being encountered by numerical modeling and simulation on nanosecond laser-target interactions and its potential development direction are considered.

physics.plasm-ph↗

arXiv2026

HERMES: a multi-agent framework for structured knowledge extraction from ultra-long documents in geoscience

Authoritative scientific knowledge in geoscience remains largely trapped in legacy monographs and historical literature, where unstructured text and complex layouts hinder computational access. We introduce HERMES, a scalable multi-agent framework that extracts structured data from ultra-long scientific documents. Using a coordinating large language model, HERMES integrates domain constraints, validation rules and evidence tracing within a unified document-level extraction process that incorporates parsed text, tables, figures and captions. Applied to the 55-volume Treatise on Invertebrate Paleontology, the system produced a structured database of 32,277 fossil taxonomic entities and 451,878 attributes, released online at https://treatise.geolex.org. Extraction performance remained stable across fossil groups (average F1 scores of approximately 0.90 for entities and 0.91 for attributes), improving per-volume efficiency approximately sixfold relative to the tested fully manual baseline. Evaluation in palaeomagnetism and geochemistry, conducted without additional model training, demonstrated transfer across distinct geoscience domains. This work provides a practical pathway to transform historical scientific literature into FAIR-oriented structured data, offering a sustainable infrastructure for data-intensive disciplines and large-scale knowledge integration.

cs.CL↗

arXiv2026

Quark matter at finite temperature and proto-quark stars with the axion effects in SU(3) Nambu-Jona-Lasinio model

We investigate the thermodynamical properties of strange quark matter (SQM) and proto-quark stars (PQSs) within the SU(3) Nambu-Jona-Lasinio (NJL) model at finite temperature, specifically incorporating the effects of axion fields and vector interactions. Our results demonstrate that these interactions significantly influence the equation of state (EoS), constituent quark masses, entropy density, and the maximum star mass of PQSs at the isentropic stages along the star evolution line. Furthermore, we reveal a distinct thermodynamic signature in the early evolution: the presence of trapped neutrinos leads to a substantial increase in electron number density while simultaneously suppressing the core temperature compared to the neutrino-free case. These findings may highlight the crucial role of the axion effects, flavor-dependent vector interactions, and particle composition in determining the observable properties of compact stars at finite temperature.

nucl-th↗

arXiv2026

Nine Raters, One Index: Carrying LLM Disagreement into Labour-Market Estimates

When a large language model supplies research annotations, the choice of model becomes an analytic degree of freedom. We ask how much that choice matters by putting 1,100 task-occupation cells to nine models from three vendors and weighting their ratings with worker-reported task importance in the British Skills and Employment Surveys. Rankings are substantially more stable than levels: pairwise rank correlations range from 0.74 to 0.92, while the share of British jobs scoring above 0.5 ranges from under 0.1% to 38%. We therefore treat the rater as one dimension of a multiverse, reporting each rater-specific estimate and its range alongside a Rubin-style summary that incorporates their observed dispersion. The consequences depend on the downstream design. A post-2022 pay gradient of -0.020 log points per interquartile range is negative under all nine raters, although it continues a pre-existing trend. In online vacancies, all nine raters initially produce negative post-2022 gradients; after teleworkability is given its own period path, seven produce significant positive coefficients, two produce small negative coefficients, and the pooled estimate is indistinguishable from zero. Neither application identifies an effect of generative AI. Validation against worker-reported AI use, competing exposure measures, assistant traffic and an independent task survey supports incremental predictive content and portability, but any bias shared across raters remains unidentified. The paper's contribution is a procedure for exposing and carrying observed LLM-rater disagreement into empirical estimates.

econ.GN↗

arXiv2026

MGDFIS: Multi-scale Global-detail Feature Integration Strategy for Small Object Detection

Small-object detection in Unmanned Aerial Vehicle (UAV) imagery requires preserving weak local evidence while using broader context to separate tiny foreground targets from cluttered backgrounds. Existing multi-scale fusion methods improve feature aggregation, but they often add computation or blur fine details during repeated cross-scale fusion. The central challenge is to balance low-SNR target preservation, clutter suppression, and efficient cross-scale context exchange. To address this challenge, we propose the Multi-scale Global-detail Feature Integration Strategy (MGDFIS), a neck-level feature-fusion strategy that couples global context exchange, local-detail recovery, and pixel-level foreground-background recalibration. MGDFIS integrates three coordinated modules: FusionLock-TSS Attention for stabilizing spectral-spatial responses, Global-detail Integration for combining long-range mixing with local detail capture, and Dynamic Pixel Attention for reweighting compact foreground regions. On the controlled VisDrone setting, YOLO26m + MGDFIS improves AP50:95 from 25.7 to 30.2 and AP50 from 37.2 to 44.2 over the YOLO26m baseline, with 96.1 GFLOPs. Additional dataset-specific evaluations report 38.9 AP50 and 21.9 AP50:95 on UAVDT and 97.4 AP50 on CARPK. The code is available at: https://github.com/Bai-Xuecheng/MGDFIS.

cs.CV↗

arXiv2026

No Single Neuron of Failure: Distributed Safety Alignment Against White-Box Attacks

With the rapid release of open-weight large foundation models, safety threats are shifting from black-box jailbreaks to neuron-level white-box attacks that directly identify and manipulate safety-related neurons. Existing alignment methods often investigate the safety behavior on a small number of neurons, creating fragile single point of failure with limited redundancy. To address this issue, we propose distributed safety alignment (DSA), which redundantly encodes safety capabilities across multiple computational neurons, ensuring that the model maintains its safety baseline even when critical safety neurons are disrupted. Specifically, we localize the intervention to the inputs of the down-projection layers in language-side feed-forward networks and treat each feature coordinate as the activation of an individual neuron. DSA then combines neuron activations with loss gradients to compute a direction-aware first-order Taylor score that globally identifies the neurons that contribute most to the current refusal behavior of the model. Finally, targeted disruption via deterministic masking and stochastic dropout is coupled, forcing the model to abandon narrow safety neurons and redundantly encode safety behavior across multiple compensatory neurons. Extensive experiments show that DSA substantially improves robustness against white-box neuron-level safety attacks while preserving the model's general language and multimodal utility.

cs.AI↗

arXiv2026

Cleavage-History-Dependent Low-Temperature ARPES Spectra of Charge-Ordered EuAl$_4$

Charge ordering in EuAl$_4$ has been widely discussed in connection with band reconstruction, magnetism, and topological electronic states, yet the microscopic origin of the complex low-temperature ARPES spectra remains unresolved. Here we combine photon-energy-, temperature-, and cleavage-history-dependent ARPES with first-principles calculations to distinguish intrinsic bulk bands from surface-preparation-dependent spectral weight. Spectra measured on high-temperature-cleaved surfaces, both at 160 K and after cooling to 10 K, are broadly consistent with the calculated three-dimensional bulk electronic structure, whereas low-temperature-cleaved surfaces exhibit additional electron-like bands, replica-like Fermi-surface contours, and a pronounced $δ$ band near -0.57 eV that is absent from the calculated bulk bands. The additional features are observed at multiple photon energies and on multiple independently cleaved surfaces and are selectively suppressed upon warming, while the bulk-derived bands remain comparatively stable. The $δ$ band does not emerge when the same high-temperature-cleaved surface is cooled through $T_{\rm CDW}$. Comparison with the projected bulk bands and the calculated spectral function of an ideal Eu-terminated surface further associates the additional bands with the surface electronic structure. These results establish a strong cleavage-history dependence of the low-temperature ARPES spectra and provide spectroscopic criteria for separating surface-reconstruction and bulk charge-order contributions in EuAl$_4$.

cond-mat.str-el↗

arXiv2026

Measurement of multiple mechanical properties from multi-dimensional signals in nanosecond laser ablation via PINN

Accurate evaluation of mechanical properties in steels under ageing or service conditions remains a major challenge. We propose a thermo-mechanical coupling framework for nanosecond laser ablation based on energy conservation, which is embedded into a physics-informed neural network (PINN) to enable simultaneous inversion of multiple mechanical properties. A thermo-mechanical coupling coefficient is defined to uniformly describe the dynamic allocation of input laser energy among thermal diffusion, mechanical work and plasma shielding across different deformation stages under laser irradiation. Furthermore, hard-to-measure physical characteristics in the coupled equation are replaced with experimentally accessible features obtained through the simultaneous acquisition of spectroscopic, shockwave and surface-wave signals. Using 210 experimental datasets, the framework simultaneously recovers Young's modulus, yield strength, ultimate tensile strength and micro-Vickers hardness with high accuracy (R2=0.9927, 0.9912, 0.9916 and 0.9959 respectively), significantly outperforming the baseline method (ultrasonic velocity regression for E, R2=0.0012). Comparisons with linear normalization and unconstrained neural networks demonstrate that PINN achieves near-unity accuracy through the embedding of conservation-law constraints. Partial dependency analysis further uncovers the nonlinear coupling laws between input features and mechanical properties. The proposed paradigm, integrating conservation laws, measurable features and physics-informed learning, offers a universal approach for non-contact, high-precision and physically consistent multi-to-multi inversion of multiple material properties under nanosecond laser ablation conditions.

physics.plasm-ph↗