arXiv subjects

Hao Zhou

Publications and source records attributed to Hao Zhou.

At least 145 records · Page 8Linked to original sources

arXiv2026

Higher Satisfaction, Lower Cost: A Technical Report on How LLMs Revolutionize Meituan's Intelligent Interaction Systems

Enhancing customer experience is essential for business success, particularly as service demands grow in scale and complexity. Generative artificial intelligence and Large Language Models (LLMs) have empowered intelligent interaction systems to deliver efficient, personalized, and 24/7 support. In practice, intelligent interaction systems encounter several challenges: (1) Constructing high-quality data for cold-start training is difficult, hindering self-evolution and raising labor costs. (2) Multi-turn dialogue performance remains suboptimal due to inadequate intent understanding, rule compliance, and solution extraction. (3) Frequent evolution of business rules affects system operability and transferability, constraining low-cost expansion and adaptability. (4) Reliance on a single LLM is insufficient in complex scenarios, where the absence of multi-agent frameworks and effective collaboration undermines process completeness and service quality. (5) The open-domain nature of multi-turn dialogues, lacking unified golden answers, hampers quantitative evaluation and continuous optimization. To address these challenges, we introduce WOWService, an intelligent interaction system tailored for industrial applications. With the integration of LLMs and multi-agent architectures, WOWService enables autonomous task management and collaborative problem-solving. Specifically, WOWService focuses on core modules including data construction, general capability enhancement, business scenario adaptation, multi-agent coordination, and automated evaluation. Currently, WOWService is deployed on the Meituan App, achieving significant gains in key metrics, e.g., User Satisfaction Metric 1 (USM 1) -27.53% and User Satisfaction Metric 2 (USM 2) +25.51%, demonstrating its effectiveness in capturing user needs and advancing personalized service.

cs.CL↗

arXiv2026

Three-body resonances of $ααM$ clusters ($M=ϕ$, $J/ψ$, $η_c$) in $^{8}_{M}{\mathrm{Be}}$ nuclei

Motivated by the recently obtained HAL QCD potentials for the $N$-$ϕ$, $N$-$J/ψ$, and $N$-$η_c$ interactions, we investigate the structure of the exotic nuclei $^{8}_ϕ{\text{Be}}$, $^{8}_{J/ψ}{\text{Be}}$, and $^{8}_{η_c}{\text{Be}}$ as $α+α+M$ three-body systems ($M$ denotes the meson). The bound and resonant states are calculated consistently using the Gaussian expansion method, with resonances identified via the complex scaling method. For the $αϕ$ and $α$-charmonium interactions, a folding potential is constructed based on the HAL QCD potentials and fitted to a Woods-Saxon form. We find that the $ϕ$ meson exhibits a strong ``glue-like" effect, binding the $0^+_1$, $2^+_1$, and $4^+_1$ resonant states of $^8$Be into stable states and significantly reducing the $α$-$α$ distance. In contrast, the interactions of $J/ψ$ and $η_c$ with the nucleus are weaker, forming only shallow bound states with the $0^+_1$ state of $^8$Be and even increasing the $α$-$α$ separation. Notably, our analysis predicts weakly bound $α$-$J/ψ$ states in the $^4S_{3/2}$ and $^2S_{1/2}$ channels, a result not reported in prior studies, which suggests that $^{8}_{J/ψ}{\text{Be}}$ may not be a Borromean nucleus. The sensitivity of the $^{8}_{M}{\mathrm{Be}}(4^+_1)$ state-transitioning from bound to resonant depending on the $α$-particle radius-highlights the subtle dynamics at play. These results provide a systematic theoretical comparison of how different vector mesons modify nuclear clustering, offering critical predictions for future experimental searches of such exotic hadron-nucleus systems.

hep-ph↗

arXiv2026

La RoSA: Enhancing LLM Efficiency via Layerwise Rotated Sparse Activation

Activation sparsity can reduce the computational overhead and memory transfers during the forward pass of Large Language Model (LLM) inference. Existing methods face limitations, either demanding time-consuming recovery training that hinders real-world adoption, or relying on empirical magnitude-based pruning, which causes fluctuating sparsity and unstable inference speed-up. This paper introduces LaRoSA (Layerwise Rotated Sparse Activation), a novel method for activation sparsification designed to improve LLM efficiency without requiring additional training or magnitude-based pruning. We leverage layerwise orthogonal rotations to transform input activations into rotated forms that are more suitable for sparsification. By employing a Top-K selection approach within the rotated activations, we achieve consistent model-level sparsity and reliable wall-clock time speed-up. LaRoSA is effective across various sizes and types of LLMs, demonstrating minimal performance degradation and robust inference acceleration. Specifically, for LLaMA2-7B at 40% sparsity, LaRoSA achieves a mere 0.17 perplexity gap with a consistent 1.30x wall-clock time speed-up, and reduces the accuracy gap in zero-shot tasks compared to the dense model to just 0.54%, while surpassing TEAL by 1.77% and CATS by 17.14%.

cs.CL↗

arXiv2026

CubeBench: Diagnosing Interactive, Long-Horizon Spatial Reasoning Under Partial Observations

Large Language Model (LLM) agents, while proficient in the digital realm, face a significant gap in physical-world deployment due to the challenge of forming and maintaining a robust spatial mental model. We identify three core cognitive challenges hindering this transition: spatial reasoning, long-horizon state tracking via mental simulation, and active exploration under partial observation. To isolate and evaluate these faculties, we introduce CubeBench, a novel generative benchmark centered on the Rubik's Cube. CubeBench uses a three-tiered diagnostic framework that progressively assesses agent capabilities, from foundational state tracking with full symbolic information to active exploration with only partial visual data. Our experiments on leading LLMs reveal critical limitations, including a uniform 0.00% pass rate on all long-horizon tasks, exposing a fundamental failure in long-term planning. We also propose a diagnostic framework to isolate these cognitive bottlenecks by providing external solver tools. By analyzing the failure modes, we provide key insights to guide the development of more physically-grounded intelligent agents.

cs.AI↗

arXiv2025

VSA:Visual-Structural Alignment for UI-to-Code

The automation of user interface development has the potential to accelerate software delivery by mitigating intensive manual implementation. Despite the advancements in Large Multimodal Models for design-to-code translation, existing methodologies predominantly yield unstructured, flat codebases that lack compatibility with component-oriented libraries such as React or Angular. Such outputs typically exhibit low cohesion and high coupling, complicating long-term maintenance. In this paper, we propose \textbf{VSA (VSA)}, a multi-stage paradigm designed to synthesize organized frontend assets through visual-structural alignment. Our approach first employs a spatial-aware transformer to reconstruct the visual input into a hierarchical tree representation. Moving beyond basic layout extraction, we integrate an algorithmic pattern-matching layer to identify recurring UI motifs and encapsulate them into modular templates. These templates are then processed via a schema-driven synthesis engine, ensuring the Large Language Model generates type-safe, prop-drilled components suitable for production environments. Experimental results indicate that our framework yields a substantial improvement in code modularity and architectural consistency over state-of-the-art benchmarks, effectively bridging the gap between raw pixels and scalable software engineering.

cs.IR↗

arXiv2025

Modular Layout Synthesis (MLS): Front-end Code via Structure Normalization and Constrained Generation

Automated front-end engineering drastically reduces development cycles and minimizes manual coding overhead. While Generative AI has shown promise in translating designs to code, current solutions often produce monolithic scripts, failing to natively support modern ecosystems like React, Vue, or Angular. Furthermore, the generated code frequently suffers from poor modularity, making it difficult to maintain. To bridge this gap, we introduce Modular Layout Synthesis (MLS), a hierarchical framework that merges visual understanding with structural normalization. Initially, a visual-semantic encoder maps the screen capture into a serialized tree topology, capturing the essential layout hierarchy. Instead of simple parsing, we apply heuristic deduplication and pattern recognition to isolate reusable blocks, creating a framework-agnostic schema. Finally, a constraint-based generation protocol guides the LLM to synthesize production-ready code with strict typing and component props. Evaluations show that MLS significantly outperforms existing baselines, ensuring superior code reusability and structural integrity across multiple frameworks

cs.IR↗

arXiv2025

Hypervelocity Impact Debris Cloud Trajectory-Planning based on Additive Manufactured Lattice Structures

Space debris and micrometeoroid (MMOD) impacts pose a serious threat to the safe operation of spacecraft. However, traditional protective structures typically suffer from limitations such as excessive thickness and inadequate load-bearing capacity. Guided by the design concepts of debris-cloud deflection and hierarchical energy dissipation, this study proposes a trajectory-planning lattice protective structure. First, the lattice parameters and geometry were designed according to the functional relationship between the incident angle and the transmitted/ricochet trajectory angles. Subsequently, multi-angle hypervelocity impact experiments were carried out to evaluate the proposed lattice protection structure. In combination with post-impact CT three-dimensional reconstruction and smoothed particle hydrodynamics (SPH) numerical simulations, the protective mechanisms of the lattice structure were systematically characterized and clarified. The results demonstrate that, for three oblique incidence conditions, the lattice structure remained intact and significantly deflected the debris-cloud momentum direction while effectively dissipating its kinetic energy. The angled plates with gradient designs enabled continuous changes in the momentum direction and stepwise kinetic energy dissipation through multiple cycles of debrisation, dispersion, and trajectory deflection. This research presents a novel, engineering-ready approach for spacecraft MMOD protection and validates the potential of trajectory-planning lattice structures for hypervelocity impact defense.

physics.app-ph↗

arXiv2025

Local discontinuous Galerkin method for the integral fractional Laplacian

We develop and analyze a local discontinuous Galerkin (LDG) method for solving integral fractional Laplacian problems on bounded Lipschitz domains. The method is based on a three-field mixed formulation involving the primal variable, its gradient, and the corresponding Riesz potential, yielding a flux-based structure well suited for LDG discretizations while retaining the intrinsic nonlocal interaction. A key ingredient of our analysis is a rigorous study of the weighted Hölder and Sobolev regularity of the Riesz potential, which enables accurate characterization of boundary singularities. Guided by these regularity results, we propose LDG schemes on quasi-uniform and graded meshes, with additional stabilization in the graded case to reconcile the discrepancy between the discrete spaces for the Riesz potential and flux fields. Optimal a priori error estimates are established, and numerical experiments corroborate the theoretical results.

math.NA↗

arXiv2025

SRLR: Symbolic Regression based Logic Recovery to Counter Programmable Logic Controller Attacks

Programmable Logic Controllers (PLCs) are critical components in Industrial Control Systems (ICSs). Their potential exposure to external world makes them susceptible to cyber-attacks. Existing detection methods against controller logic attacks use either specification-based or learnt models. However, specification-based models require experts' manual efforts or access to PLC's source code, while machine learning-based models often fall short of providing explanation for their decisions. We design SRLR -- a it Symbolic Regression based Logic Recovery} solution to identify the logic of a PLC based only on its inputs and outputs. The recovered logic is used to generate explainable rules for detecting controller logic attacks. SRLR enhances the latest deep symbolic regression methods using the following ICS-specific properties: (1) some important ICS control logic is best represented in frequency domain rather than time domain; (2) an ICS controller can operate in multiple modes, each using different logic, where mode switches usually do not happen frequently; (3) a robust controller usually filters out outlier inputs as ICS sensor data can be noisy; and (4) with the above factors captured, the degree of complexity of the formulas is reduced, making effective search possible. Thanks to these enhancements, SRLR consistently outperforms all existing methods in a variety of ICS settings that we evaluate. In terms of the recovery accuracy, SRLR's gain can be as high as 39% in some challenging environment. We also evaluate SRLR on a distribution grid containing hundreds of voltage regulators, demonstrating its stability in handling large-scale, complex systems with varied configurations.

cs.LG↗

arXiv2025

MACS: Multi-source Audio-to-image Generation with Contextual Significance and Semantic Alignment

Propelled by the breakthrough in deep generative models, audio-to-image generation has emerged as a pivotal cross-modal task that converts complex auditory signals into rich visual representations. However, previous works only focus on single-source audio inputs for image generation, ignoring the multi-source characteristic in natural auditory scenes, thus limiting the performance in generating comprehensive visual content. To bridge this gap, we propose a method called MACS to conduct multi-source audio-to-image generation. To our best knowledge, this is the first work that explicitly separates multi-source audio to capture the rich audio components before image generation. MACS is a two-stage method. In the first stage, multi-source audio inputs are separated by a weakly supervised method, where the audio and text labels are semantically aligned by casting into a common space using the large pre-trained CLAP model. We introduce a ranking loss to consider the contextual significance of the separated audio signals. In the second stage, effective image generation is achieved by mapping the separated audio signals to the generation condition using only a trainable adapter and a MLP layer. We preprocess the LLP dataset as the first full multi-source audio-to-image generation benchmark. The experiments are conducted on multi-source, mixed-source, and single-source audio-to-image generation tasks. The proposed MACS outperforms the current state-of-the-art methods in 17 out of the 21 evaluation indexes on all tasks and delivers superior visual quality.

cs.SD↗

arXiv2025

The Formation Rate and Luminosity Function of Fast X-ray transients from Einstein probe

Following its launch on 2024 January 9, the Einstein Probe (EP) telescope has detected hundreds of fast X-ray transients (FXTs), yet their physical origins remain elusive. Understanding their luminosity function and formation rate is crucial for elucidating their nature. Recently, the EP team has provided the latest catalog of EP-detected FXTs. Based on this catalog, we present a model-independent nonparametric approach to derive the luminosity function and formation rate of FXTs. Our analysis reveals significant cosmological luminosity evolution, characterized by a scaling relationship of $(1+z)^{3.58}$. After accounting for this evolution, we establish that the local luminosity function is best represented by a broken power law, with a break luminosity of $(4.17 \pm 0.34) \times 10^{46}$ erg/s. The formation rate exhibits a broken power law as $ρ(z) \propto (1+z)^{-4.25}$ at $z \lessapprox 0.9$ and $ρ(z) \propto (1+z)^{-0.26}$ at $z \gtrapprox 0.9$, yielding a local rate of approximately $153.8_{-95.1}^{+249.4}$ Gpc$^{-3}$ yr$^{-1}$. This rate is higher than that of long gamma-ray bursts (LGRBs). Our findings indicate that a component of FXTs is associated with LGRBs.

astro-ph.HE↗

arXiv2025

FLEX: Continuous Agent Evolution via Forward Learning from Experience

Autonomous agents driven by Large Language Models (LLMs) have revolutionized reasoning and problem-solving but remain static after training, unable to grow with experience as intelligent beings do during deployment. We introduce Forward Learning with EXperience (FLEX), a gradient-free learning paradigm that enables LLM agents to continuously evolve through accumulated experience. Specifically, FLEX cultivates scalable and inheritable evolution by constructing a structured experience library through continual reflection on successes and failures during interaction with the environment. FLEX delivers substantial improvements on mathematical reasoning, chemical retrosynthesis, and protein fitness prediction (up to 23% on AIME25, 10% on USPTO50k, and 14% on ProteinGym). We further identify a clear scaling law of experiential growth and the phenomenon of experience inheritance across agents, marking a step toward scalable and inheritable continuous agent evolution. Project Page: https://flex-gensi-thuair.github.io.

cs.LG↗

arXiv2025

EP241217a: a likely Type II GRB with an achromatic bump at z = 4.59

EP241217a is an X-ray transient detected by the Einstein Probe (EP) lasting for about 100 seconds and without accompanying $γ$-ray detection. The optical spectroscopy reveals the redshift of EP241217a is 4.59. By combining the $γ$-ray upper limit provided by GECAM-C, there is a considerable possibility that EP241217a is a typical Type II gamma-ray burst (GRB), but it is fainter than the detection threshold of any available $γ$-ray monitors (i.e., $E_{γ,{\rm iso}}\lesssim10^{53}$ erg). The X-ray light curve exhibits a plateau lasting for $\sim5\times10^4$ seconds. However, the joint analysis with optical data suggests the presence of an achromatic bump peaking at $\sim3\times10^4$ s after the trigger, indicating the actual duration of the X-ray plateau may be significantly shorter than it appears. To interpret the achromatic bump, we adopt the scenario of a mildly relativistic jet coasting in a wind-like medium and encountering a rapid density enhancement of the circumburst medium, which is likely induced by the the interaction of the progenitor's stellar wind and the interstellar medium. However, this model cannot fully explain observed data, and some issues do exist, e.g., the observed spectrum is harder than the model prediction. Consequently, we conclude that the scenario of a mildly relativistic jet coasting in the wind-like medium cannot explain all observed features of EP241217a. In addition, some alternative models commonly invoked to explain X-ray plateaus are discussed, but there are more or less issues when they are applied to EP241217a. Therefore, further theoretical modeling is encouraged to explore the origin of EP241217a.

astro-ph.HE↗

arXiv2025

FiCoTS: Fine-to-Coarse LLM-Enhanced Hierarchical Cross-Modality Interaction for Time Series Forecasting

Time series forecasting is central to data analysis and web technologies. The recent success of Large Language Models (LLMs) offers significant potential for this field, especially from the cross-modality aspect. Most methods adopt an LLM-as-Predictor paradigm, using LLM as the forecasting backbone and designing modality alignment mechanisms to enable LLM to understand time series data. However, the semantic information in the two modalities of time series and text differs significantly, making it challenging for LLM to fully understand time series data. To mitigate this challenge, our work follows an LLM-as-Enhancer paradigm to fully utilize the advantage of LLM in text understanding, where LLM is only used to encode text modality to complement time series modality. Based on this paradigm, we propose FiCoTS, an LLM-enhanced fine-to-coarse framework for multimodal time series forecasting. Specifically, the framework facilitates progressive cross-modality interaction by three levels in a fine-to-coarse scheme: First, in the token-level modality alignment module, a dynamic heterogeneous graph is constructed to filter noise and align time series patches with text tokens; Second, in the feature-level modality interaction module, a global cross-attention mechanism is introduced to enable each time series variable to connect with relevant textual contexts; Third, in the decision-level modality fusion module, we design a gated network to adaptively fuse the results of the two modalities for robust predictions. These three modules work synergistically to let the two modalities interact comprehensively across three semantic levels, enabling textual information to effectively support temporal prediction. Extensive experiments on seven real-world benchmarks demonstrate that our model achieves state-of-the-art performance. The codes will be released publicly.

cs.LG↗

arXiv2025

Self-Refined Generative Foundation Models for Wireless Traffic Prediction

With a broad range of emerging applications in 6G networks, wireless traffic prediction has become a critical component of network management. However, the dynamically shifting distribution of wireless traffic in non-stationary 6G networks presents significant challenges to achieving accurate and stable predictions. Motivated by recent advancements in Generative AI (GenAI)-enabled 6G networks, this paper proposes a novel self-refined Large Language Model (LLM) for wireless traffic prediction, namely TrafficLLM, through in-context learning without parameter fine-tuning or model training. The proposed TrafficLLM harnesses the powerful few-shot learning abilities of LLMs to enhance the scalability of traffic prediction in dynamically changing wireless environments. Specifically, our proposed TrafficLLM embraces an LLM to iteratively refine its predictions through a three-step process: traffic prediction, feedback generation, and prediction refinement. Initially, the proposed TrafficLLM conducts traffic predictions using task-specific demonstration prompts. Recognizing that LLMs may generate incorrect predictions on the first attempt, this paper designs feedback demonstration prompts to provide multifaceted and valuable feedback related to these initial predictions. The validation scheme is further incorporated to systematically enhance the accuracy of mathematical calculations during the feedback generation process. Following this comprehensive feedback, our proposed TrafficLLM introduces refinement demonstration prompts, enabling the same LLM to further refine its predictions and thereby enhance prediction performance. Evaluations on two realistic datasets demonstrate that the proposed TrafficLLM outperforms LLM-based in-context learning methods, achieving performance improvements of 23.17% and 17.09%, respectively.

eess.SY↗

arXiv2025

Conan: Progressive Learning to Reason Like a Detective over Multi-Scale Visual Evidence

Video reasoning, which requires multi-step deduction across frames, remains a major challenge for multimodal large language models (MLLMs). While reinforcement learning (RL)-based methods enhance reasoning capabilities, they often rely on text-only chains that yield ungrounded or hallucinated conclusions. Conversely, frame-retrieval approaches introduce visual grounding, yet still struggle with inaccurate evidence localization. To address these limitations, we present Conan, a framework for evidence-grounded multi-step video reasoning. Conan identifies context and evidence frames, reasons over cross-frame clues, and adaptively decides when to conclude or explore further. To achieve this, we 1) construct Conan-91K, a large-scale dataset of automatically generated reasoning traces that include frame identification, evidence reasoning, and action decision, and 2) design a multi-stage progressive cold-start strategy combined with an Identification-Reasoning-Action (AIR) RLVR training framework to progressively incentivize multi-step visual reasoning. Extensive experiments on six multi-step reasoning benchmarks demonstrate that Conan surpasses the baseline Qwen2.5-VL-7B-Instruct by an average of over 10% in accuracy, achieving state-of-the-art performance. Furthermore, Conan generalizes effectively to long video understanding tasks, validating its strong scalability and robustness.

cs.CV↗

arXiv2025

Credit Network Modeling and Analysis via Large Language Models

We investigate the application of large language models (LLMs) to construct credit networks from firms' textual financial statements and to analyze the resulting network structures. We start with using LLMs to translate each firm's financial statement into a credit network that pertains solely to that firm. These networks are then aggregated to form a comprehensive credit network representing the whole financial system. During this process, the inconsistencies in financial statements are automatically detected and human intervention is involved. We demonstrate that this translation process is effective across financial statements corresponding to credit networks with diverse topological structures. We further investigate the reasoning capabilities of LLMs in analyzing credit networks and determining optimal strategies for executing financial operations to maximize network performance measured by the total assets of firms, which is an inherently combinatorial optimization challenge. To demonstrate this capability, we focus on two financial operations: portfolio compression and debt removal, applying them to both synthetic and real-world datasets. Our findings show that LLMs can generate coherent reasoning and recommend effective executions of these operations to enhance overall network performance.

cs.MA↗

arXiv2025

HiMAE: Hierarchical Masked Autoencoders Discover Resolution-Specific Structure in Wearable Time Series

Wearable sensors provide abundant physiological time series, yet the principles governing their predictive utility remain unclear. We hypothesize that temporal resolution is a fundamental axis of representation learning, with different clinical and behavioral outcomes relying on structure at distinct scales. To test this resolution hypothesis, we introduce HiMAE (Hierarchical Masked Autoencoder), a self supervised framework that combines masked autoencoding with a hierarchical convolutional encoder decoder. HiMAE produces multi resolution embeddings that enable systematic evaluation of which temporal scales carry predictive signal, transforming resolution from a hyperparameter into a probe for interpretability. Across classification, regression, and generative benchmarks, HiMAE consistently outperforms state of the art foundation models that collapse scale, while being orders of magnitude smaller. HiMAE is an efficient representation learner compact enough to run entirely on watch, achieving sub millisecond inference on smartwatch class CPUs for true edge inference. Together, these contributions position HiMAE as both an efficient self supervised learning method and a discovery tool for scale sensitive structure in wearable health.

cs.LG↗