arXiv subjects

Tao Peng

Publications and source records attributed to Tao Peng.

At least 19 recordsLinked to original sources

arXiv2026

Radiation, Rotation and Scale Invariant Feature Descriptor for Multimodal Image Matching

Multimodal image matching is a fundamental task for multi-source information fusion. However, geometric distortions and nonlinear radiometric differences (NRD) severely limit performance, especially under radiometric, rotation, and scale variations. To address this issue, we propose a radiation, rotation, and scale invariant (RRSI) feature descriptor. First, a dual-head regional sampling (DHRS) module simultaneously performs Cartesian and Log-Polar sampling on keypoint neighborhoods, retaining spatial structural properties while enhancing robustness to rotation and scale variations. We then jointly encode geometric and radiometric relations between multimodal images in a unified deep feature space, enabling feature encoding, interaction, and fusion across intra-modal, dual-head sampled, and inter-modal regions. Furthermore, we introduce a bidirectional cross-modal generative reconstruction constraint during training. By decoding implicit features into structural patches of the counterpart modality, this mechanism anchors modality-invariant geometric topologies without additional inference overhead. Experiments on optical-infrared and optical-SAR datasets demonstrate highly competitive matching performance and strong robustness to rotation and scale variations. RRSI supports the full rotation range from 0 to 360 degrees and scale factors up to four. Its generalization ability is further validated on multimodal images from computer vision, remote sensing, and medical imaging. The implementation will be made publicly available at https://github.com/yeyuanxin110/RRSI .

cs.CV↗

arXiv2026

Rethinking AI Cloud Infrastructure for Agentic Serving Systems with the Aries Experimentation Framework

Autonomous agents challenge conventional LLM serving by coupling repeated inference with persistent context and sandboxed tool execution. We present Aries, a full-stack experimentation framework that separates task semantics from execution configurations, reconstructs cross-component agent trajectories with correlated system telemetry, and exposes stateful tool execution through a consistent interface across heterogeneous sandbox substrates. We use Aries to conduct reproducible experiments on open agent harnesses and benchmarks. We complement these experiments with production traces from a commercial platform, grounding low-level systems research in observed production behavior. Our results show that (1) token-centric metrics miss non-inference bottlenecks, (2) retaining additional context yields diminishing accuracy benefits while reducing serving capacity, and (3) tool sandboxes alternate between long idle periods and short resource bursts, while current snapshot-based state management makes aggressive suspension costly. A complementary security analysis further highlights the need to reduce the sandbox attack surface. We then discuss the vision for agent-native serving systems designed around trajectory-level metrics, adaptive context management, elastic sandbox resource management, and sandboxes with minimized attack surface.

cs.DC↗

arXiv2026

Towards Precision Therapy in Hepatocellular Carcinoma: A Clinical-Reasoning LLM for Risk Stratification and Treatment Guidance

Hepatocellular carcinoma (HCC) is a common malignancy and a leading cause of cancer-related mortality. Current guidelines and staging systems provide coarse categories, but often miss within-stage heterogeneity and the clinical context in electronic medical records (EMRs). We present HCC-STAR (Hepatocellular Carcinoma Staging, Treatment And pRognosis), a clinically aligned large language model that reads routine EMR narratives and jointly outputs risk score-based staging, ranked guideline-consistent treatments with evidence-based rationales, and individualized survival estimates. We curated about 30,000 HCC cases from SEER and expanded them into EMR-style narrative training data using a clinician-validated, prompt-based augmentation workflow. On this corpus, we developed a knowledge-aligned reasoning framework optimized with a step-verifiable composite reward, moving beyond text-level memorization of clinical guidelines. In a multi-center cohort of 6,668 patients from 12 hospitals in China, HCC-STAR achieved state-of-the-art performance in treatment recommendation and risk stratification compared with clinical guidelines and competitive models, including GPT-5 and Gemini-2.5 Pro. Hypothetical overall-survival analysis showed a median survival of 51 months under adherence to HCC-STAR recommendations, compared with 29 and 32 months under BCLC and CNLC. In clinician-centric evaluations, blinded hepatobiliary specialists rated HCC-STAR's reasoning and evidence-based justifications as trustworthy. The model surpassed resident and attending physicians in treatment accuracy and helped physicians make more accurate decisions faster when used as an assistant. These findings support HCC-STAR as a reliable and verifiable decision-support system for risk stratification and precision therapy in HCC.

cs.AI↗

arXiv2026

FAME: Forecastability-Aware Mixture of Experts for Heterogeneous Time Series Forecasting

Large-scale retail and industrial forecasting systems contain many heterogeneous time series whose lifecycle, sparsity, volatility, seasonality, spectral patterns, and contextual sensitivity differ substantially. A single forecasting model rarely performs well across all regimes, while dense ensembles increase inference cost and provide limited insight into expert suitability. This paper studies forecastability-aware expert routing: learning how data characteristics determine the suitability of forecasting experts. We propose \method{}, a sparse mixture-of-experts framework that represents each series with a multidimensional forecastability fingerprint, mines expert-suitability targets from validation performance, and trains a cost-aware sparse router to activate a small budgeted set of experts for each series. Using a production-scale vending-machine sales dataset from Shandong New Beiyang (SNBC), where the forecasting component has been integrated into the replenishment-planning pipeline, together with public retail benchmarks, we show that expert suitability varies systematically across data regimes. On the industrial dataset with 5,000+ machines and 60M+ transactions, \method{} Top-2 reduces MSE by 12.4\% over the strongest single expert, LightGBM, while executing 1.92 experts per series on average. The deployed component produces demand forecasts, while inventory-oriented gains are estimated by an offline replay simulator under a fixed replenishment policy rather than by online intervention. The framework turns heterogeneous sales forecasting from heuristic model selection into data mining of forecastability patterns and expert specialization. Code is available at https://github.com/hit636/FAME

cs.AI↗

arXiv2026

Development of a high-field pulsed magnet and optical fiber coupled cryostat system for magneto-photoluminescence measurements

This paper presents the development of a high field pulsed magnet system combined with an optical fiber-coupled low temperature cryostat for magneto-photoluminescence measurements. A novel aspect of our system is the use of electrolyte capacitors in a 75 kJ capacitor bank to drive the pulsed magnet, enabling the generation of high magnetic fields at a relatively low charging voltage (400 V). The zylon reinforced wire wound magnet coil achieves a maximum field strength of 35 tesla with a magnetic field rise time of 10 ms. The integrated cryostat was developed to meet the specific requirements of magneto-photoluminescence experiments using a 4 K helium closed-cycle cryocooler to provide a stable low-temperature environment down to 5 K.

physics.ins-det↗

arXiv2026

\$OneMillion-Bench: How Far are Language Agents from Human Experts?

As language models (LMs) evolve from chat assistants to long-horizon agents capable of multi-step reasoning and tool use, existing benchmarks remain largely confined to structured or exam-style tasks that fall short of real-world professional demands. To this end, we introduce \$OneMillion-Bench \$OneMillion-Bench, a benchmark of 400 expert-curated tasks spanning Law, Finance, Industry, Healthcare, and Natural Science, built to evaluate agents across economically consequential scenarios. Unlike prior work, the benchmark requires retrieving authoritative sources, resolving conflicting evidence, applying domain-specific rules, and making constraint decisions, where correctness depends as much on the reasoning process as the final answer. We adopt a rubric-based evaluation protocol scoring factual accuracy, logical coherence, practical feasibility, and professional compliance, focused on expert-level problems to ensure meaningful differentiation across agents. Together, \$OneMillion-Bench provides a unified testbed for assessing agentic reliability, professional depth, and practical readiness in domain-intensive scenarios.

cs.LG↗

arXiv2026

Experimental investigation of wall-pressure fluctuations on a fully appended submarine model at high Reynolds numbers

This paper addresses a critical gap in hydroacoustics through a systematic wind tunnel investigation of wall-pressure fluctuations on the fully appended DARPA SUBOFF model at operationally relevant Reynolds numbers ranging from $5.6 \times 10^{6}$ to $1.4 \times 10^{7}$. The experimental campaign encompasses baseline straight-ahead flow, complex maneuvering (yaw and pitch) conditions, and a first-of-its-kind assessment of a novel vortex control baffle (VCB). To ensure benchmark-quality spectral data, rigorous signal processing techniques were applied, specifically Wiener filtering for background noise suppression and dynamic transfer function correction for pinhole sensors. Key findings indicate that while spectral self-similarity holds across Reynolds numbers, the primary finding is the critical role of appendages in noise amplification. Unstable horseshoe vortex dynamics at the sail-hull junction drive localized pressure fluctuations of up to 300%, establishing this feature as a major coherent noise source. To address this, the study provides the pioneering experimental validation of the VCB. By physically suppressing horseshoe vortex formation at the sail-hull junction, the VCB achieves a global stabilization of the downstream flow, resulting in a significant 35% reduction in root-mean-square wall-pressure fluctuations at the stern and an approximately 14% reduction along the parallel mid-body. Furthermore, maneuvering conditions are shown to fundamentally reshape the pressure field, introducing substantial crossflow effects and non-monotonic spectral behaviors. This comprehensive dataset and the demonstrated efficacy of the VCB provide essential physical insights and a critical validation benchmark for the design of next-generation quiet submarines.

physics.flu-dyn↗

arXiv2026

The Impact of Interference Cognition on the Reliability and Capacity of Industrial Wireless Communications

Interference significantly impacts the performance of industrial wireless networks, particularly n severe interference environments with dense networks reusing spectrum resources intensively. Although delicate interference information is often unavailable in conventional networks, emerging interference cognition techniques can compensate this critical problem with possibly different precision. This paper investigates the relationship between precision of interference cognition and system performance. We propose a novel performance analysis framework that quantifies the impact of varying interference information precision on achievable rate. Specifically, leveraging the Nakagami-$\mathbf{m}$ fading channel model, we analytically and asymptotically analyze the average achievable rate in the finite blocklength regime for different precision levels of signal and interference information. Our findings reveal the critical importance of identifying per-link interference information for achieving optimal performance. Additionally, obtaining instantaneous information is more beneficial for signal links.

eess.SY↗

arXiv2026

Efficient Code Analysis via Graph Representation Learning-Guided Large Language Models

Large Language Models (LLMs) have significantly advanced code analysis tasks, yet they struggle to detect malicious behaviors fragmented across files, whose intricate dependencies easily get lost in the vast amount of benign code. We therefore propose a graph-centric attention acquisition pipeline that enhances LLMs' ability to localize malicious behavior. The approach parses a project into a code graph, uses an LLM to encode nodes with semantic and structural signals, and trains a Graph Neural Network (GNN) under sparse supervision. The GNN performs an initial detection, and by interpreting these predictions, identifies key code sections that are most likely to contain malicious behavior. These influential regions are then used to guide the LLM's attention for in-depth analysis. This strategy significantly reduces interference from irrelevant context while maintaining low annotation costs. Extensive experiments show that the method consistently outperforms existing approaches on multiple public and custom datasets, highlighting its potential for practical deployment in software security scenarios.

cs.SE↗

arXiv2025

Experimental study on the wall-pressure fluctuations of flow over an axisymmetric hull

Wall pressure fluctuations beneath the turbulent boundary layer of high-speed underwater vehicles are crucial for hydro-acoustics and acoustic stealth. However, a comprehensive understanding remains limited due to a lack of high-quality experimental data, particularly under realistic operational conditions. To address this gap, this study establishes the first high-fidelity experimental database of wall-pressure fluctuations on an axisymmetric hull at high Reynolds numbers. The dataset's primary innovation is its systematic inclusion of complex maneuvering (yaw and pitch) conditions, providing a benchmark for validating flow noise prediction models. Analysis of this dataset yields key physical insights. The study quantifies systematic Reynolds number effects, including a spectral energy shift toward lower frequencies, and spectral scaling laws by revealing the critical influence of pressure-gradient effects. These findings provide fundamental insights into non-equilibrium 3D turbulent flows and establish an essential dataset to support the design of quieter and more effective underwater vehicles.

physics.flu-dyn↗

arXiv2025

ProServe: Unified Multi-Priority Request Scheduling for LLM Serving

The widespread deployment of large language models (LLMs) for interactive applications necessitates serving systems that can handle thousands of concurrent requests with diverse Service Level Objective (SLO) requirements. A critical yet often overlooked dimension in this context is the inherent priority difference among clients; for instance, business-critical functions demand higher performance guarantees, as fulfilling such requests yields significantly greater business value. However, existing LLM serving schedulers fail to jointly optimize for both SLO attainment and client-level priorities. To bridge this gap, we first formalize multi-priority request scheduling as a service gain maximization problem, where satisfying latency requirements for requests of different priorities contributes varying gain. We propose ProServe, a unified two-tier scheduling framework designed to maximize overall service gain. At the engine layer, SlideBatching dynamically adapts batch formation under varying loads, employing a sliding boundary mechanism to balance latency and priority differentiation. Considering potential preemption, efficient block management adopts asynchronous offloading, pipelined reloading, and adaptive copy-budget control to overlap computation with host-device block transfers. At the service layer, GoRouting performs gain-oriented and capability-aware dispatching across distributed instances, proactively reserving capacity for future high-priority or long requests. Extensive evaluation on four open-source and one industrial dataset shows that ProServe outperforms state-of-the-art baselines, improving system gain by up to 35% and SLO attainment by up to 52%.

cs.DC↗

arXiv2025

xLLM Technical Report

We introduce xLLM, an intelligent and efficient Large Language Model (LLM) inference framework designed for high-performance, large-scale enterprise-grade serving, with deep optimizations for diverse AI accelerators. To address these challenges, xLLM builds a novel decoupled service-engine architecture. At the service layer, xLLM-Service features an intelligent scheduling module that efficiently processes multimodal requests and co-locates online and offline tasks through unified elastic scheduling to maximize cluster utilization. This module also relies on a workload-adaptive dynamic Prefill-Decode (PD) disaggregation policy and a novel Encode-Prefill-Decode (EPD) disaggregation policy designed for multimodal inputs. Furthermore, it incorporates a distributed architecture to provide global KV Cache management and robust fault-tolerant capabilities for high availability. At the engine layer, xLLM-Engine co-optimizes system and algorithm designs to fully saturate computing resources. This is achieved through comprehensive multi-layer execution pipeline optimizations, an adaptive graph mode and an xTensor memory management. xLLM-Engine also further integrates algorithmic enhancements such as optimized speculative decoding and dynamic EPLB, collectively serving to substantially boost throughput and inference efficiency. Extensive evaluations demonstrate that xLLM delivers significantly superior performance and resource efficiency. Under identical TPOT constraints, xLLM achieves throughput up to 1.7x that of MindIE and 2.2x that of vLLM-Ascend with Qwen-series models, while maintaining an average throughput of 1.7x that of MindIE with Deepseek-series models. xLLM framework is publicly available at https://github.com/jd-opensource/xllm and https://github.com/jd-opensource/xllm-service.

cs.DC↗

arXiv2025

CryptoX : Compositional Reasoning Evaluation of Large Language Models

The compositional reasoning capacity has long been regarded as critical to the generalization and intelligence emergence of large language models LLMs. However, despite numerous reasoning-related benchmarks, the compositional reasoning capacity of LLMs is rarely studied or quantified in the existing benchmarks. In this paper, we introduce CryptoX, an evaluation framework that, for the first time, combines existing benchmarks and cryptographic, to quantify the compositional reasoning capacity of LLMs. Building upon CryptoX, we construct CryptoBench, which integrates these principles into several benchmarks for systematic evaluation. We conduct detailed experiments on widely used open-source and closed-source LLMs using CryptoBench, revealing a huge gap between open-source and closed-source LLMs. We further conduct thorough mechanical interpretability experiments to reveal the inner mechanism of LLMs' compositional reasoning, involving subproblem decomposition, subproblem inference, and summarizing subproblem conclusions. Through analysis based on CryptoBench, we highlight the value of independently studying compositional reasoning and emphasize the need to enhance the compositional reasoning capabilities of LLMs.

cs.CR↗

arXiv2025

UI-TARS: Pioneering Automated GUI Interaction with Native Agents

This paper introduces UI-TARS, a native GUI agent model that solely perceives the screenshots as input and performs human-like interactions (e.g., keyboard and mouse operations). Unlike prevailing agent frameworks that depend on heavily wrapped commercial models (e.g., GPT-4o) with expert-crafted prompts and workflows, UI-TARS is an end-to-end model that outperforms these sophisticated frameworks. Experiments demonstrate its superior performance: UI-TARS achieves SOTA performance in 10+ GUI agent benchmarks evaluating perception, grounding, and GUI task execution. Notably, in the OSWorld benchmark, UI-TARS achieves scores of 24.6 with 50 steps and 22.7 with 15 steps, outperforming Claude (22.0 and 14.9 respectively). In AndroidWorld, UI-TARS achieves 46.6, surpassing GPT-4o (34.5). UI-TARS incorporates several key innovations: (1) Enhanced Perception: leveraging a large-scale dataset of GUI screenshots for context-aware understanding of UI elements and precise captioning; (2) Unified Action Modeling, which standardizes actions into a unified space across platforms and achieves precise grounding and interaction through large-scale action traces; (3) System-2 Reasoning, which incorporates deliberate reasoning into multi-step decision making, involving multiple reasoning patterns such as task decomposition, reflection thinking, milestone recognition, etc. (4) Iterative Training with Reflective Online Traces, which addresses the data bottleneck by automatically collecting, filtering, and reflectively refining new interaction traces on hundreds of virtual machines. Through iterative training and reflection tuning, UI-TARS continuously learns from its mistakes and adapts to unforeseen situations with minimal human intervention. We also analyze the evolution path of GUI agents to guide the further development of this domain.

cs.AI↗

arXiv2024

SDM-Car: A Dataset for Small and Dim Moving Vehicles Detection in Satellite Videos

Vehicle detection and tracking in satellite video is essential in remote sensing (RS) applications. However, upon the statistical analysis of existing datasets, we find that the dim vehicles with low radiation intensity and limited contrast against the background are rarely annotated, which leads to the poor effect of existing approaches in detecting moving vehicles under low radiation conditions. In this paper, we address the challenge by building a \textbf{S}mall and \textbf{D}im \textbf{M}oving Cars (SDM-Car) dataset with a multitude of annotations for dim vehicles in satellite videos, which is collected by the Luojia 3-01 satellite and comprises 99 high-quality videos. Furthermore, we propose a method based on image enhancement and attention mechanisms to improve the detection accuracy of dim vehicles, serving as a benchmark for evaluating the dataset. Finally, we assess the performance of several representative methods on SDM-Car and present insightful findings. The dataset is openly available at https://github.com/TanedaM/SDM-Car.

cs.CV↗

arXiv2024

DefFiller: Mask-Conditioned Diffusion for Salient Steel Surface Defect Generation

Current saliency-based defect detection methods show promise in industrial settings, but the unpredictability of defects in steel production environments complicates dataset creation, hampering model performance. Existing data augmentation approaches using generative models often require pixel-level annotations, which are time-consuming and resource-intensive. To address this, we introduce DefFiller, a mask-conditioned defect generation method that leverages a layout-to-image diffusion model. DefFiller generates defect samples paired with mask conditions, eliminating the need for pixel-level annotations and enabling direct use in model training. We also develop an evaluation framework to assess the quality of generated samples and their impact on detection performance. Experimental results on the SD-Saliency-900 dataset demonstrate that DefFiller produces high-quality defect images that accurately match the provided mask conditions, significantly enhancing the performance of saliency-based defect detection models trained on the augmented dataset.

cs.CV↗

arXiv2024

Elastic properties of Cu-6wt\%Ag alloy wires for pulsed magnets investigated by ultrasonic techniques

Conductor materials with good mechanical performance as well as high electrical- and thermal-conductivities are particularly important to break through the current bottle-neck limit ($\sim 100$ T) of pulsed magnets. Here we perform systematic studies on the elastic properties of the Cu-6wt\%Ag alloy wires, a promising candidate material for the new-generation pulsed magnets, by employing two independent ultrasonic techniques - resonant ultrasound spectroscopy (RUS) and ultrasound pulse-echo experiments. Our RUS measurements manifest that the elastic properties of the Cu-6wt\%Ag alloy wires can be improved by an electroplastic drawing procedure as compared with the conventional cold drawing. We also take this chance to test the availability of our newly-built ultrasound pulse-echo facility at Wuhan National High Magnetic Field Center (WHMFC, China), and the results suggest that the elastic performance of the electroplastically-drawn Cu-6wt\%Ag alloy wire remains excellent without anomalous softening under extreme conditions, e.g., ultra-high magnetic field up to 50 T, nitrogen / helium cryogenic liquids.

cond-mat.mtrl-sci↗

arXiv2024

An Empirical Study on Information Extraction using Large Language Models

Human-like large language models (LLMs), especially the most powerful and popular ones in OpenAI's GPT family, have proven to be very helpful for many natural language processing (NLP) related tasks. Therefore, various attempts have been made to apply LLMs to information extraction (IE), which is a fundamental NLP task that involves extracting information from unstructured plain text. To demonstrate the latest representative progress in LLMs' information extraction ability, we assess the information extraction ability of GPT-4 (the latest version of GPT at the time of writing this paper) from four perspectives: Performance, Evaluation Criteria, Robustness, and Error Types. Our results suggest a visible performance gap between GPT-4 and state-of-the-art (SOTA) IE methods. To alleviate this problem, considering the LLMs' human-like characteristics, we propose and analyze the effects of a series of simple prompt-based methods, which can be generalized to other LLMs and NLP tasks. Rich experiments show our methods' effectiveness and some of their remaining issues in improving GPT-4's information extraction ability.

cs.CL↗