Search arXiv⌕ Search

arXiv subjects

Yang Yu

Publications and source records attributed to Yang Yu.

At least 37 records · Page 2Linked to original sources
arXiv2026

OpenHarmony Bench: Evaluating LLMs and Coding Agents on OpenHarmony App Development

We present OPENHARMONY BENCH, an app-level coding benchmark for evaluating LLM-based coding agents on OpenHarmony ArkTS applications. Unlike function-level benchmarks, it evaluates complete app-level changes: each task requires an agent to modify a buildable ArkTS project so that a requested behavior works end to end, involving UI state, data persistence, build configuration, and platform APIs. The benchmark installs and drives the delivered application on a device to check whether the behavior is observable. It covers three input sources: natural-language feature requests (new-feature), structured scenario specifications (spec-driven), and bug descriptions (bug-fix). The benchmark contains 153 top-level tasks and 242 Feature points (F-points), where an F-point is one executable behavior check. The snapshot includes 32 new-feature tasks, 50 spec-driven tasks with 139 F-points, and 71 bug-fix tasks. The main leaderboard is scored over top-level tasks rather than independently weighted F-points. We describe the benchmark construction, statistics, and build-and-test evaluation pipeline, and evaluate DevEco Code with eight LLMs across three independent full-suite runs per configuration. Three findings emerge. First, newer generations complete more tasks than their predecessors within evaluated model-family pairs. Second, buildability is close to saturated while behavioral correctness is not: mean Final Build Success Rate is 94.77% to 100.00%, whereas mean Task Completion is 48.36% to 58.39%. Third, spec-driven tasks have the lowest Task Completion under all-checks task scoring, with no configuration exceeding 35%. The code, data, tasks, reference solutions, tests, evaluation scripts, and leaderboard are released through the official OPENHARMONY BENCH website at https://bench.matrix.openharmony.cn/.

cs.SE↗
arXiv2026

EA-LiteUNet: An Edge-Adaptive and Resource-Efficient U-Net for Boundary-Sensitive Dermoscopic Image Segmentation

Accurate boundary delineation remains a persistent challenge in dermoscopic image segmentation because of blurred lesion margins, heterogeneous textures, and complex background artifacts. From a signal-processing perspective, lesion boundaries represent high-frequency components that are highly susceptible to aliasing, noise amplification, and information loss. Consequently, repeated downsampling and feature transformations in conventional convolutional architectures often lead to severely degraded boundary representations. To address these limitations, we propose EA-LiteUNet, an edge-adaptive and computationally efficient U-Net variant specifically designed for boundary-sensitive medical image segmentation. The architecture integrates three core mechanisms: (1) boundary-aware representation learning to suppress aliasing and preserve high-frequency structural details; (2) attention-guided feature modulation to selectively enhance boundary-relevant responses across multi-scale features; and (3) a resource-adaptive inference strategy to dynamically balance segmentation accuracy and computational efficiency. Extensive evaluations across three public dermoscopic datasets demonstrate that EA-LiteUNet consistently achieves superior boundary precision. Specifically, on the ISIC 2018 dataset, the method significantly reduces the 95% Hausdorff Distance (HD95) to 12.89 pixels while maintaining a robust Dice score of 92.08%. Notably, this strong performance is achieved with an ultralightweight configuration of merely 0.29M parameters and 1.17 GFLOPs. Ablation studies further validate the complementary effects of these components, confirming their contribution to enhanced boundary fidelity and stable optimization.

cs.CV↗
arXiv2026

A scalable edge-pass Purcell filter for high-fidelity readout of superconducting qubits

High-fidelity readout with strong Purcell protection of qubit coherence is essential for scalable superconducting quantum processors, yet the finite passband and sizable footprint of conventional band-pass Purcell filters make them hard to scale. Here we introduce a scalable edge-pass Purcell filter that separates the readout band from the protected qubit band by a single transmission edge, freeing the readout resonators from bandwidth constraint. Depending on whether the transmitting band lies above or below the cutoff, the compact network is realized as a high-pass filter (HPF) or a low-pass filter (LPF). The HPF reaches an average readout fidelity of 99.46(4)% (up to 99.56%) with a 150-ns pulse, and the LPF reaches 99.49(3)% (up to 99.57%) with a 130-ns pulse. The average single-qubit gate fidelities are 99.94% (HPF) and 99.93% (LPF). Relative to the filter-free Purcell limit, the filters substantially extend the qubit lifetime, and the Purcell protection deepens at higher filter order. In addition, an intrinsic dissipation mode of the filter offers a qubit-reset channel. This leads to a compact architecture that unifies fast, high-fidelity readout, Purcell protection, and effective reset within a single filter for large-scale fault-tolerant quantum computation.

quant-ph↗
arXiv2026

Properties of holographic superconductors from Machine Learning

We investigate holographic superconductors using modern optimisation techniques inspired by machine learning. The critical temperature is obtained by minimising the variational functional for the eigenvalue $λ^2$ with two complementary trial functions: a simple cosine ansatz $F(z)=\cos(a z)$ and a flexible exponential polynomial $F(z)=\exp(\sum_{n=2}^{N} a_n z^n)$, both of which automatically satisfy the standard boundary conditions. For the cosine ansatz, we perform a one-parameter minimisation and obtain $λ^2(Δ)$ and $T_c/\sqrtρ$ over a wide range of $Δ$, including the exact values at $Δ=1$ and $Δ=2$ to high accuracy. The exponential polynomial ansatz, with up to 19 coefficients, is optimised using a multi-start L-BFGS-B algorithm with warm-starting, yielding even better agreement with known exact results. Our numerical data for $λ^2(Δ)$ and $T_c/\sqrtρ$ match the analytical predictions from the literature, confirming the robustness of the variational approach. This work; therefore, demonstrates that a combination of analytic trial functions and modern numerical optimisation provides a powerful, flexible, and efficient tool for exploring holographic superconductors, and can be readily extended to include backreaction or other sectors in this field.

hep-th↗
arXiv2026

Synchrotron self-Compton process for constraining sub-GeV dark matter in Omega Centauri via SKA

The search for the particle identity of dark matter (DM) continues to be a primary objective in modern physics. In this field, the sub-GeV mass range of DM detection remains a crucial yet challenging window. We investigate synchrotron self-Compton (SSC) emission from electrons and positrons produced by MeV-scale DM annihilation as a novel indirect detection channel. Focusing on the globular cluster Omega Centauri and the sensitivity of the Square Kilometre Array, we derive constraints on the annihilation cross section reaching $\langleσv\rangle \sim 10^{-29}\,\rm{cm}^{3}\,\rm{s}^{-1}$ in the tens-of-MeV range. Furthermore, constraints can reach below $\langleσv\rangle \sim 10^{-30}\,\rm{cm}^{3}\,\rm{s}^{-1}$ under favorable parameter choices. Although the derived limits depend on the uncertain propagation parameters, the SSC channel remains competitive with existing indirect constraints over a representative range of astrophysical assumptions, establishing SSC emission as a promising probe of sub-GeV DM.

astro-ph.CO↗
arXiv2026

ELBench: A Multi-Dimensional Benchmark for Education-Facing Large Language Models

Large language models are increasingly deployed in education as tutors, teaching assistants, and content generators. These roles place demands that ordinary question answering does not: a usable education-facing model is supposed to be accurate, safe under sensitive prompts, instructionally useful, and aligned with pedagogical goals at the same time. Existing benchmarks evaluate these requirements largely in isolation, so none assesses education-facing suitability as an integrated profile. We introduce ELBench, the first benchmark to evaluate all four requirements (General Capability, Safety and Trustworthiness, Basic Education, and High-Level Cultivation) on the same models under a common protocol, combining curated public sources with newly synthesized safety and cultivation data. We evaluate nine models, seven frontier general-purpose systems and two education-specialized variants, and report three findings. First, module-level profiles are more informative than a single aggregate: the top six models are statistically indistinguishable on overall score, yet their module leaders differ substantially, and safety is anti-correlated with practical teaching (r = -0.83). Second, the Chinese-developed models lead the safety module, the most discriminative in the suite; this advantage is largest on region-specific normative content and narrows, but does not vanish, on universal-harm content. Third, the two education-specialized models lead neither education module, and on High-Level Cultivation all models share a systematic blind spot: on the structured judgment task they converge on the same non-reference option, favoring pedagogical style over fit to the stated goal, so the module scores uniformly low and does not separate models. This raises, but does not resolve, whether domain post-training keeps pace with frontier systems on education tasks.

cs.CL↗
arXiv2026

Direct experimental measurement of femtonewton-scale momentum transfer force from electron beams

Electron beams (e-beams) are ubiquitous in imaging, patterning, and propulsion. This prevalence is rooted in the profound mastery of their wave-particle duality and energy-transfer pathways. Yet, a fundamental dimension remains largely unexplored: while the mechanical effect (i.e., the momentum transfer to a target) is theoretically known, quantification of its femtonewton-range force has remained elusive. This discrepancy represents a missing piece of the puzzle toward a comprehensive understanding of e-beam-matter interactions, and ultimately limits the multi-dimensional exploitation of e-beams. A force sensor combining femtonewton sensitivity, immunity to electromagnetic noise, compatibility with vacuum, and absolute calibration is critical to bridge the gap between theory and experiment. Here the FINEST (Femtonewton Interferometric Nanomechanical Electron-beam Sensing Technology) sensor is proposed and successfully tested to measure the force of an e-beam. FINEST is an optical-pressure-calibrated 3D spring-type optical sensor that operates reliably under e-beam conditions. Femtonewton-scale forces from 2-30 keV e-beams are directly measured, ranging from 505 fN to 13 pN. Both linear scaling with beam current and a non-monotonic energy dependence (peaking near 10 keV) are observed. Based on this calibrated force, the mechanical contribution to e-beam ice etching was quantitatively confirmed; its effect is orders of magnitude lower than the total etch depth and lacks noticeable energy dependence. By achieving the first direct experimental measurement of e-beam momentum transfer, this work adds a long-missing dimension to the physical landscape of e-beam processes. These findings provide a quantitative basis for furthering the multi-dimensional exploitation of e-beams, potentially transforming our approach to precision nanofabrication, sensing, and fundamental electron physics research.

physics.optics↗
arXiv2026

Potential Matching Optimal Transport: Continuous Normalizing Flows for Exact $p$-Wasserstein Dynamics

We introduce Potential Matching Optimal Transport (PMOT), a potential-flow framework for general $p$-cost optimal transport with $c_p(x,y)=\|x-y\|^p$. PMOT parameterizes the CNF velocity field with a scalar potential in the generalized Benamou--Brenier form for the chosen exponent $p$. It trains the potential gradient with a self-induced matching loss along straight bridges determined by the model's own endpoints, while allowing flexible terminal distribution matching. Our main result establishes zero-loss exactness: under the stated regularity, exact terminal matching, and uniqueness assumptions, any zero-loss solution satisfies the generalized Benamou--Brenier optimality system and recovers the corresponding $p$-optimal transport map and dynamics. On synthetic benchmarks, PMOT learns $p$-specific maps that agree with the corresponding $p$-matched OT references. It also remains competitive as a likelihood-based density model on high-dimensional tabular data, and an MMD-based color transformation experiment demonstrates flexible sample-based terminal matching.

cs.LG↗
arXiv2026

PDD-RRG: Posterior Diagnostic Decision for Study-level Radiology Report Generation

Automatic radiology report generation (RRG) aims to simulate the workflow of radiologists, assisting them in clinical diagnosis. However, existing methods often fall short in utilizing all information relevant to the examination, as is typically done in clinical practice. Although some works attempt to incorporate multi-view images and historical data, these additional inputs may sometimes lead to avoidable diagnostic errors on the contrary. To address these challenges, we introduce a decision-making stage after report generation for the first time and propose a Posterior Diagnostic Decision framework (PDD-RRG) to integrate potentially conflicting diagnoses. Specifically, we create various subsets of input data and utilize an existing RRG model to generate reports from different perspectives. Then the Bayesian posterior probability and the learned thresholds for each clinical observation are calculated to obtain an aggregated diagnostic conclusion, which is subsequently used to refine the generated report. Experiments on MIMIC-CXR demonstrate that our proposed PDD-RRG can effectively enhance the clinical efficacy of existing RRG models without any retraining.

cs.CV↗
arXiv2026

Beyond Captions: Context-Grounded Reconstruction for Biomedical Multimodal Continued Pretraining

Biomedical figures are explained not by captions alone but by body-text passages that discuss them. Yet current multimodal corpora typically reduce figures to isolated image-caption pairs, discarding this crucial context. Existing pipelines either omit this context or append it without enforcing the figure references that support each attachment, which can create unsupported image-text attachments and incoherent discourse. We introduce context-grounded reconstruction, a source-grounded framework that converts PubMed Central Open Access (PMC-OA) records into referentially coherent interleaved sequences. It recovers captions and source text, attaches context only through article-native figure references, repairs non-contiguous context, and prunes unsupported images. Starting from these reconstructed sequences, PMC-InterCPT first filters records for text quality and medical relevance, then applies evidence-aware allocation to form a 9.63B-token corpus for continued pretraining (CPT) of generative medical MLLMs. With fixed supervised fine-tuning (SFT), PMC-InterCPT improves Qwen3.5-4B-Base by 1.46 medical-average points and 3.11 general/scientific-average points over a token-matched raw source control, and surpasses a 42% larger raw-data run. Gains transfer to Qwen3.5-2B-Base and LLaVA-OneVision-1.5-4B-Base. Controlled ablations show that context-grounded reconstruction, rather than simply appending article context or scaling raw data, is central to useful biomedical multimodal CPT.

cs.CL↗
arXiv2026

Local Universality and Structural Certificates for Minimal Fixed-Depth Two-Qutrit Gate Decomposition

We study a dimension-saturating fixed-core ansatz in which four copies of a fixed, non-tunable two-qutrit core $K\in SU(9)$ are interleaved with five adjustable local layers from $L=SU(3)\otimes SU(3)$. Since $\dim SU(9)=80$ and $5\dim L=80$, this is the shortest fixed-core architecture not excluded by parameter counting. We formulate the smooth map $Φ_K:L^5\to SU(9)$ and use its right-trivialized differential to give verifiable certificates for local universality. We construct an explicit Clifford-word core whose Pauli-label splitting makes the identity-point differential an exact isometry, and we classify all 2304 symplectic actions satisfying the same splitting criterion. We also prove a structural obstruction for an important symmetry class: every complex-symmetric core $K=K^{T}$, including every core generated by a time-independent real-symmetric Hamiltonian in the chosen computational basis, has identity-point differential rank at most 78; hence any full-rank certificate for such a core must occur away from that point. We then assess a hardware-motivated superconducting core generated by a noncommuting, temporally asymmetric drive. Direct calculation verifies $K_{\rm sc}\neq K_{\rm sc}^{\mathsf T}$, and the core achieves $F_{\rm avg}\ge 0.999$ for all 1000 Haar-random targets tested under the stated restart protocol. We also report favorable sampled Jacobian-rank, structured-target, and robustness diagnostics. These results establish local universality at the parameter-counting-minimal, dimension-saturating depth, with an exact Clifford certificate complemented by a hardware-motivated numerical case study. Throughout, we separate exact local certificates from numerical evidence for broader synthesis performance.

quant-ph↗
arXiv2026

Large Language Models Versus Physicians in Traditional Chinese Medicine: A Real-World Clinical Case Evaluation

Large language models (LLMs) are increasingly being explored for clinical applications, yet their assessment for real-world traditional Chinese medicine (TCM) practice remains limited We constructed a clinical case library comprising 349 de-identified outpatient cases from 62 hospitals and evaluated 16 LLMs and a comparator cohort of 60 practicing TCM physicians using 60 representative cases selected from this library. Model outputs and physician reports were anonymized and scored by five senior TCM experts across nine diagnostic and therapeutic dimensions. Cutting-edge general-purpose LLMs achieved higher expert scores than the physician comparators, particularly for medical advice, treatment principles and selected diagnostic tasks. However, prescription-level analyses revealed discrepancies in herb selection, dosage, and treatment strategy, and qualitative safety review identified hallucinations and undesirable template-driven outputs. These findings highlight the potential of LLMs for TCM decision support while underscoring the need for physician oversight, safety constraints and prospective clinical evaluation.

cs.CL↗
arXiv2026

Efficient Test-Time Optimization for Multi-Agent Proof Autoformalization

Full-proof autoformalization bridges extensive mathematical proofs in natural language with formally validated reasoning, offering a pathway to elevate the ceiling of verifiable mathematical reasoning. Unlike statement-level formalization, proof autoformalization is a long-horizon challenge requiring coordination of claims, contexts, and dependencies across many proof steps, yet has only recently come under focused study. Current approaches either rely on costly model training or apply excessive, unguided repair at inference time. To this end, we introduce ToMap, a multi-agent framework that structures proof autoformalization as a Decomposer-Formalizer-Prover pipeline with efficient test-time optimization guided by formal verification and semantic rubrics for proof quality. Rather than distributing test-time compute across all agents, we perform bottleneck analysis and identify the Decomposer as the critical bottleneck: the quality of its atomic, self-contained proof units directly determines whether downstream agents can successfully formalize and prove each step. ToMap therefore treats the Formalizer and Prover as downstream executors and efficiently focuses test-time compute on Decomposer refinement. This refinement follows a loop inspired by GEPA, evolving prompts over candidate decompositions and using formal verification progress together with semantic proof rubrics to define a Pareto frontier that guides the next decomposition update. Experiments on ProofFlowBench show that ToMap improves over the best previous method by 19.0% when evaluated by both syntactic correctness and semantic faithfulness, while requiring lower test-time cost. Scaling analysis shows that most gains emerge within a few iterations of decomposition evolution, guiding test-time budget selection.

cs.AI↗
arXiv2026

Event Stream based Multi-Modal Video Anomaly Detection: A Benchmark Dataset and Algorithms

Video anomaly detection (VAD) is critical for automated surveillance but remains fragile under challenging conditions such as illumination variations, fast motion, and complex backgrounds when relying solely on visible light videos. To address these limitations, we propose EVAD, an event enhanced VAD framework that jointly exploits conventional video and event streams captured by bio inspired event cameras. Event sensors asynchronously capture brightness changes with high temporal resolution, offering robustness to motion blur and extreme lighting, and providing motion salient cues complementary to video based visual information. To support multi modal VAD research, we construct a large scale visible event benchmark comprising 6.3 billion events and 376,368 video frames collected under diverse illumination levels, motion patterns, and background complexities, filling the gap of realistic and scalable datasets for event based anomaly detection. Building upon this dataset, we design a contrastive multi modal pretraining framework to learn discriminative event representations by aligning semantic embeddings across event streams, visible videos, and textual descriptions. An adaptive fusion module then dynamically integrates event based temporal cues with video based spatial semantics, improving robustness to environmental disturbances. Experiments on benchmarks and the proposed TJUTCM Pha dataset demonstrate that E VAD consistently outperforms methods, validating the effectiveness of event-based sensing for VAD in real world scenarios.

cs.CV↗
arXiv2026

Hidden Decoding at Scale: Latent Computation Scaling for Large Language Models

Scaling Large Language Models (LLMs) has been driven mainly by enlarging the Transformer backbone, but for an already-strong model this requires another round of costly pretraining. We study whether an existing backbone can keep improving by allocating more computation to each token while leaving the Transformer backbone fixed. Depth-recurrent (looped) Transformers pursue this goal but are hard to scale, because looped computation does not fit naturally with the pipeline parallelism used to train the largest models. We add computation along the sequence-length dimension, where the extra computation is simply a longer input and stays compatible with standard large-model training. We propose Hidden Decoding, a sequence-length scaling method applied during continued pretraining (CPT). It expands each token into n streams with independent embedding tables and keeps the intermediate streams' key-value cache as context, so each token performs more internal computation without adding or widening Transformer layers. To keep this affordable at scale, we introduce Stream-Factorized Attention, in which most layers attend only within each stream and only a few layers mix across streams, reducing the attention cost from quadratic to roughly linear in n. Experiments support two scaling results. At frontier scale, we train WeLM-HD4-80B and WeLM-HD4-617B at n=4 and improve their matched non-HD baselines, making Hidden Decoding the first demonstrated sequence-length scaling method at the 100B+ MoE scale. Across expansion factors, the gains grow as n increases, showing that sequence-length expansion is a practical fixed-backbone scaling path for frontier-scale LLMs.

cs.CL↗
arXiv2026

AI Adoption in S&P 500 Firms

The adoption of artificial intelligence (AI) by large enterprises is an important potential source of aggregate productivity improvement and labor market impact. We study AI adoption of S&P 500 firms over the period 2016 to 2025, estimating adoption at the enterprise level. While generative AI tools are useful for personal and professional applications, our focus is on the deep integration of AI in the business processes of large enterprises which are bellwethers for firm adoption more broadly. We develop a novel measure to assess deep AI adoption (and distinguish it from AI hype) that is based on SEC 10-K filings, where laws and regulations ``prohibit companies from making materially false or misleading statements." In 2025, 11% of S&P 500 enterprises had AI deeply integrated into their business processes, and a further 10% were using AI in the production of goods and delivery of services. AI adoption has more than quadrupled from 5% in 2022 with slowly accelerating adoption among non-technology firms but very aggressive adoption in the technology sector which accounts for two-thirds of deeply integrated enterprise adoption. Firm profitability shows a "J-curve" as firms move from no adoption to deep adoption, but we observe no differences in capex or productivity. Among technology firms, but not others, AI adoption is higher for firms with more employees and higher values of Tobin's q.

econ.GN↗
arXiv2026

Seeing Through Multiple Views: Parameter-Efficient Fine-Tuning via Selective Neurons for Consistent Radiology Report Generation

Recent years have seen substantial advances in radiology report generation (RRG), yet existing approaches predominantly adopt direct feature fusion when handling multi-view X-ray images. Such approaches overlook the potential clinical inconsistencies and inaccuracies arising when a single model processes different views, adversely impacting performance and clinical reliability. To this end, we introduce View-PNDF (View-specific Pattern Neuron Detection and Fine-tuning), a parameter-efficient framework that fosters view-consistent report generation from a neuronal perspective. Specifically, View-PNDF comprises: (i) a view-specific neuron detection module identifying neurons responsive to particular views, (ii) a verification module quantifying the existence of these neurons, and (iii) a selective fine-tuning strategy strengthening detected neurons while preserving view-agnostic representations. By updating only view-specific neurons, View-PNDF achieves consistent diagnoses across different views with reduced computational costs. Subsequently, we employ Large Language Models (LLMs) to consolidate the view-specific reports into a complete radiology report. Furthermore, we use traditional Natural Language Generation (NLG) metrics-based assessment on integrated reports for baseline comparison and employ LLM-based assessment (e.g., GPT-4o) on view-specific reports to capture clinical significance. Extensive experiments on two medical RRG benchmarks demonstrate that View-PNDF substantially improves view-specific chest X-ray report generation quality while maintaining robust general-view performance.

cs.CV↗
arXiv2026

How Should World Models Be Evaluated for Embodied Decision-Making? A Decision-Making-Centric Position

World models have become a central abstraction in modern AI. The term now refers to several different objects: action-conditioned environment models, latent imagination models, future-video predictors, interactive neural simulators, latent predictive representations, and synthetic-data engines. Evaluation has broadened along with the term. Recent papers measure video realism, perceptual similarity, instruction following, physical plausibility, policy ranking, executability, planning success, and downstream policy improvement. This produces both metric diversity and a recurring problem of claim/evidence mismatch: papers sometimes make a stronger claim about what their model is useful for than their evaluation can establish. This paper surveys the recent literature and argues that, for models presented as world models for embodied decision-making, the more decisive issue is not whether the model generates visually convincing videos, but whether it supports reliable interventional reasoning, policy evaluation, planning, and policy optimization under intervention, policy-induced distribution shift, and long-horizon rollout. We organize the survey using an L0--L7 ladder spanning visual plausibility to policy optimization utility, noting that the levels cut across several orthogonal axes and so form an evidential hierarchy rather than a single scalar. The framework foregrounds interventional action fidelity, closed-loop rollout validity, reward/value prediction, policy-ranking agreement, optimization lift, model exploitability, and uncertainty calibration, with a minimal feasible reporting set for real-robot settings.

cs.LG↗