arXiv subjects

Zhihao Wang

Publications and source records attributed to Zhihao Wang.

At least 37 records · Page 2Linked to original sources

arXiv2026

The Potential of Erroneous Outbound Traffic Analysis to Unveil Silent Internal Anomalies

Passive measurement has traditionally focused on inbound traffic to detect malicious activity, based on the assumption that threats originate externally. In this paper, we offer a complementary perspective by examining outbound traffic, and argue that a narrow subset -- what we term erroneous outbound traffic -- is a lighter and revealing yet overlooked data source for identifying a broad range of security threats and network problems. This traffic consists of packets sent by internal hosts that either receive no response, trigger ICMP errors, or are ICMP error messages themselves generated in response to unsolicited requests. To demonstrate its potential, we collect and analyse erroneous traffic from a large network, uncovering a variety of previously unnoticed issues, including misconfigurations, obsolete deployments and compromised hosts.

cs.NI↗

arXiv2026

SAFE-QAQ: End-to-End Slow-Thinking Audio-Text Fraud Detection via Reinforcement Learning

Existing fraud detection methods predominantly rely on transcribed text, suffering from ASR errors and missing crucial acoustic cues like vocal tone and environmental context. This limits their effectiveness against complex deceptive strategies. To address these challenges, we first propose \textbf{SAFE-QAQ}, an end-to-end comprehensive framework for audio-based slow-thinking fraud detection. First, the SAFE-QAQ framework eliminates the impact of transcription errors on detection performance. Secondly, we propose rule-based slow-thinking reward mechanisms that systematically guide the system to identify fraud-indicative patterns by accurately capturing fine-grained audio details, through hierarchical reasoning processes. Besides, our framework introduces a dynamic risk assessment framework during live calls, enabling early detection and prevention of fraud. Experiments on the TeleAntiFraud-Bench demonstrate that SAFE-QAQ achieves dramatic improvements over existing methods in multiple key dimensions, including accuracy, inference efficiency, and real-time processing capabilities. Currently deployed and analyzing over 70,000 calls daily, SAFE-QAQ effectively automates complex fraud detection, reducing human workload and financial losses. Code: https://anonymous.4open.science/r/SAFE-QAQ.

cs.SD↗

arXiv2025

Developmental trajectories of decision making and affective dynamics in large language models

Large language models (LLMs) are increasingly used in medicine and clinical workflows, yet we know little about their decision and affective profiles. Taking a historically informed outlook on the future, we treated successive OpenAI models as an evolving lineage and compared them with humans in a gambling task with repeated happiness ratings. Computational analyses showed that some aspects became more human-like: newer models took more risks and displayed more human-like patterns of Pavlovian approach and avoidance. At the same time, distinctly non-human signatures emerged: loss aversion dropped below neutral levels, choices became more deterministic than in humans, affective decay increased across versions and exceeded human levels, and baseline mood remained chronically higher than in humans. These "developmental" trajectories reveal an emerging psychology of machines and have direct implications for AI ethics and for thinking about how LLMs might be integrated into clinical decision support and other high-stakes domains.

cs.CY↗

arXiv2025

A Network Arena for Benchmarking AI Agents on Network Troubleshooting

Agentic systems, powered by Large Language Models (LLMs), assist network engineers with network configuration synthesis and network troubleshooting tasks. For network troubleshooting, progress is hindered by the absence of standardized and accessible benchmarks for evaluating LLM agents in dynamic network settings at low operational effort. We present NIKA, the largest public benchmark to date for LLM-driven network incident diagnosis and troubleshooting. NIKA targets both domain experts and especially AI researchers alike, providing zero-effort replay of real-world network scenarios, and establishing well-defined agent-network interfaces for quick agent prototyping. NIKA comprises hundreds of curated network incidents, spanning five network scenarios, from data centers to ISP networks, and covers 54 representative network issues. Lastly, NIKA is modular and extensible by design, offering APIs to facilitate the integration of new network scenarios and failure cases. We evaluate state-of-the-art LLM agents on NIKA and find that while larger models succeed more often in detecting network issues, they still struggle to localize faults and identify root causes. NIKA is open-source and available to the community: https://github.com/sands-lab/nika.

cs.NI↗

arXiv2025

K-frames: Scene-Driven Any-k Keyframe Selection for long video understanding

Multimodal Large Language Models (MLLMs) have demonstrated significant capabilities in image understanding, but long-video are constrained by context windows and computational cost. Uniform frame sampling often leads to substantial information loss. Meanwhile existing keyframe selection methods such as text-frame retrieval or RL-based frame optimization typically yield sparse and temporally disjointed frames, overlooking scene continuity and lacking flexibility for multi-scale frame selection. To address these limitations, we introduce K-frames, a novel paradigm for scene-driven keyframe selection that preserves temporal continuity. Instead of selecting individual frames, K-frames predicts semantically coherent, query-relevant clips, which enables any-k keyframes selection to meet diverse user budgets. To achieve this approach, we first introduce PeakClips, a dataset of 200K video highlights conditioned by query. Building on this dataset, K-frames learns clip2frame selection using a three-stage progressive curriculum. It involves two Supervised Fine-Tuning stages for temporal grounding and key-clip perception, followed by a Reinforcement Learning stage that directly optimizes the scene-driven prediction policy for downstream task without further annotations. Extensive experiments on major long-video understanding benchmarks demonstrate that K-frames provides an effective, interpretable, and plug-and-play solution for keyframe selection at various scales. Our dataset and model will be available.

cs.LG↗

arXiv2025

X-VLA: Soft-Prompted Transformer as Scalable Cross-Embodiment Vision-Language-Action Model

Successful generalist Vision-Language-Action (VLA) models rely on effective training across diverse robotic platforms with large-scale, cross-embodiment, heterogeneous datasets. To facilitate and leverage the heterogeneity in rich, diverse robotic data sources, we propose a novel Soft Prompt approach with minimally added parameters, by infusing prompt learning concepts into cross-embodiment robot learning and introducing separate sets of learnable embeddings for each distinct data source. These embeddings serve as embodiment-specific prompts, which in unity empower VLA models with effective exploitation of varying cross-embodiment features. Our new X-VLA, a neat flow-matching-based VLA architecture, relies exclusively on soft-prompted standard Transformer encoders, enjoying both scalability and simplicity. Evaluated across 6 simulations as well as 3 real-world robots, our 0.9B instantiation-X-VLA-0.9B simultaneously achieves SOTA performance over a sweep of benchmarks, demonstrating superior results on a wide axes of capabilities, from flexible dexterity to quick adaptation across embodiments, environments, and tasks. Website: https://thu-air-dream.github.io/X-VLA/

cs.RO↗

arXiv2025

Centers and representations of ${\rm SL}_n$ quantum Teichmüller spaces

In this paper, we compute the center of the balanced Fock-Goncharov algebra and determine its rank over the center when the quantum parameter is a root of unity. These results have potential applications to the study of the center and rank of the ${\rm SL}_n$-skein algebra. Building on this computation, we classify the irreducible representations of the balanced Fock-Goncharov algebra. Due to the Frobenius homomorphism, every irreducible representation of the (projected) ${\rm SL}_n$-skein algebra of a punctured surface $\mathfrak{S}$ determines a point in the ${\rm SL}_n$ character variety of $\mathfrak{S}$, known as the classical shadow of the representation. By pulling back the irreducible representations of the balanced Fock-Goncharov algebra via the quantum trace map, we show that there exists a ``large'' subset of the ${\rm SL}_n$ character variety such that, for any point in this subset, there exists an irreducible representation of the (projected) ${\rm SL}_n$-skein algebra whose classical shadow is this point. Finally, we prove that, under mild conditions, the representations of the ${\rm SL}_n$-skein algebra obtained in this way are independent of the choice of ideal triangulation.

math.QA↗

arXiv2025

Quantum cluster realization for projected stated ${\rm SL}_n$-skein algebras

We introduce a quantum cluster algebra structure $\mathscr A_ω(\mathfrak{S})$ inside the skew-field fractions ${\rm Frac}\bigl(\widetilde{\mathscr{S}}_ω(\mathfrak{S})\bigr)$ of the projected stated ${\rm SL}_n$-skein algebra $\widetilde{\mathscr{S}}_ω(\mathfrak{S})$ (the quotient of the reduced stated ${\rm SL}_n$-skein algebra by the kernel of the quantum trace map) for any triangulable pb surface $\mathfrak{S}$ without interior punctures. To study the relationships among the projected ${\rm SL}_n$-skein algebra $\widetilde{\mathscr{S}}_ω(\mathfrak{S})$, the quantum cluster algebra $\mathscr A_ω(\mathfrak{S})$, and its quantum upper cluster algebra $\mathscr U_ω(\mathfrak{S})$, we construct a splitting homomorphism for $\mathscr U_ω(\mathfrak{S})$ and show that it is compatible with the splitting homomorphism for $\widetilde{\mathscr{S}}_ω(\mathfrak{S})$. When every connected component of $\mathfrak{S}$ contains at least two punctures, this compatibility allows us to prove that $\widetilde{\mathscr{S}}_ω(\mathfrak{S})$ embeds into $\mathscr A_ω(\mathfrak{S})$ by showing that the stated arcs joining two distinct boundary components of $\mathfrak{S}$ (which generate $\widetilde{\mathscr{S}}_ω(\mathfrak{S})$) are, up to multiplication by a Laurent monomial in the frozen variables, exchangeable cluster variables. We further conjecture that these exchangeable cluster variables generate the quantum upper cluster algebra $\mathscr U_ω(\mathfrak{S})$, which, if true, would imply the equality $\widetilde{\mathscr{S}}_ω(\mathfrak{S})=\mathscr A_ω(\mathfrak{S})=\mathscr U_ω(\mathfrak{S})$.

math.QA↗

arXiv2025

PhysiAgent: An Embodied Agent Framework in Physical World

Vision-Language-Action (VLA) models have achieved notable success but often struggle with limited generalizations. To address this, integrating generalized Vision-Language Models (VLMs) as assistants to VLAs has emerged as a popular solution. However, current approaches often combine these models in rigid, sequential structures: using VLMs primarily for high-level scene understanding and task planning, and VLAs merely as executors of lower-level actions, leading to ineffective collaboration and poor grounding challenges. In this paper, we propose an embodied agent framework, PhysiAgent, tailored to operate effectively in physical environments. By incorporating monitor, memory, self-reflection mechanisms, and lightweight off-the-shelf toolboxes, PhysiAgent offers an autonomous scaffolding framework to prompt VLMs to organize different components based on real-time proficiency feedback from VLAs to maximally exploit VLAs' capabilities. Experimental results demonstrate significant improvements in task-solving performance on complex real-world robotic tasks, showcasing effective self-regulation of VLMs, coherent tool collaboration, and adaptive evolution of the framework during execution. PhysiAgent makes practical and pioneering efforts to integrate VLMs and VLAs, effectively grounding embodied agent frameworks in real-world settings.

cs.RO↗

arXiv2025

Center of stated $\mathrm{SL}(n)$-skein algebras

In the paper, we show some properties of (reduced) stated $\mathrm{SL}(n)$-skein algebras related to their centers for essentially bordered pb surfaces, especially their centers, finitely generation over their centers, and their PI-degrees. The proofs are based on the quantum trace maps, embeddings of (reduced) stated $\mathrm{SL}(n)$-skein algebras into quantum tori appearing in higher Teichmüller theory. Thanks to the Unicity theorem in [BG02, FKBL19], we can understand the representation theory of (reduced) stated $\mathrm{SL}(n)$-skein algebras. Moreover, the applications are beyond low-dimensional topology. For example, we can access to the representation theory of unrestricted quantum moduli algebras, and that of quantum higher cluster algebras potentially.

math.QA↗

arXiv2025

A Multi-Agent Framework with Automated Decision Rule Optimization for Cross-Domain Misinformation Detection

Misinformation spans various domains, but detection methods trained on specific domains often perform poorly when applied to others. With the rapid development of Large Language Models (LLMs), researchers have begun to utilize LLMs for cross-domain misinformation detection. However, existing LLM-based methods often fail to adequately analyze news in the target domain, limiting their detection capabilities. More importantly, these methods typically rely on manually designed decision rules, which are limited by domain knowledge and expert experience, thus limiting the generalizability of decision rules to different domains. To address these issues, we propose a MultiAgent Framework for cross-domain misinformation detection with Automated Decision Rule Optimization (MARO). Under this framework, we first employs multiple expert agents to analyze target-domain news. Subsequently, we introduce a question-reflection mechanism that guides expert agents to facilitate higherquality analysis. Furthermore, we propose a decision rule optimization approach based on carefully-designed cross-domain validation tasks to iteratively enhance the effectiveness of decision rules in different domains. Experimental results and in-depth analysis on commonlyused datasets demonstrate that MARO achieves significant improvements over existing methods.

cs.AI↗

arXiv2025

Center of stated $\mathrm{SL}(n)$-skein algebras: even roots of unity

We describe the center of quantum tori appearing in quantum higher Teichmüller theory when the quantum parameter is an even root of unity. This is a subsequent of the work in the odd roots of unity case. The centers of the quantum tori help to understand those of (reduced) stated $\mathrm{SL}(n)$-skein algebras via quantum trace maps. Moreover, we compute the PI-degree of the quantum tori, which are the same with those of (reduced) stated $\mathrm{SL}(n)$-skein algebras. As corollaries, we give matrix decompositions for anti-symmetric integer matrices for the quantum tori.

math.GT↗

arXiv2025

An Empirical Study of Federated Prompt Learning for Vision Language Model

The Vision Language Model (VLM) excels in aligning vision and language representations, and prompt learning has emerged as a key technique for adapting such models to downstream tasks. However, the application of prompt learning with VLM in federated learning (FL) scenarios remains underexplored. This paper systematically investigates the behavioral differences between language prompt learning (LPT) and vision prompt learning (VPT) under data heterogeneity challenges, including label skew and domain shift. We conduct extensive experiments to evaluate the impact of various FL and prompt configurations, such as client scale, aggregation strategies, and prompt length, to assess the robustness of Federated Prompt Learning (FPL). Furthermore, we explore strategies for enhancing prompt learning in complex scenarios where label skew and domain shift coexist, including leveraging both prompt types when computational resources allow. Our findings offer practical insights into optimizing prompt learning in federated settings, contributing to the broader deployment of VLMs in privacy-preserving environments.

cs.LG↗

arXiv2025

Dynamic Vulnerability Patching for Heterogeneous Embedded Systems Using Stack Frame Reconstruction

Existing dynamic vulnerability patching techniques are not well-suited for embedded devices, especially mission-critical ones such as medical equipment, as they have limited computational power and memory but uninterrupted service requirements. Those devices often lack sufficient idle memory for dynamic patching, and the diverse architectures of embedded systems further complicate the creation of patch triggers that are compatible across various system kernels and hardware platforms. To address these challenges, we propose a hot patching framework called StackPatch that facilitates patch development based on stack frame reconstruction. StackPatch introduces different triggering strategies to update programs stored in memory units. We leverage the exception-handling mechanisms commonly available in embedded processors to enhance StackPatch's adaptability across different processor architectures for control flow redirection. We evaluated StackPatch on embedded devices featuring three major microcontroller (MCU) architectures: ARM, RISC-V, and Xtensa. In the experiments, we used StackPatch to successfully fix 102 publicly disclosed vulnerabilities in real-time operating systems (RTOS). We applied patching to medical devices, soft programmable logic controllers (PLCs), and network services, with StackPatch consistently completing each vulnerability remediation in less than 260 MCU clock cycles.

cs.CR↗

arXiv2025

Clean few-cycle blue soliton self-compressed pulses generation in hollow-core fibers

Blue pulses with few-cycle temporal durations hold significant value in attosecond science and ultrafast spectroscopy. In this work, we combine efficient broadband frequency doubling, multiplate continuum (MPC) post-compression and blue soliton self-compression in hollow-core fibers (HCF), experimentally demonstrating HCF-based 4.4 fs clean blue soliton self-compressed pulse. Our scheme offers three-fold advantages: (1) prevention of excessive dispersion accumulation by gradually suppressing dispersion through multi-stage design; (2) enhanced overall efficiency for self-compression-based ultrashort blue pulse generation; and (3) elimination of dispersion compensation components for the final spectral broadening stage compared to post-compression schemes. This work extends HCF-based self-compression soliton to the blue spectral region and paves the way for generating energetic ultrashort blue pulses.

physics.optics↗

arXiv2025

A Novel Evaluation Benchmark for Medical LLMs: Illuminating Safety and Effectiveness in Clinical Domains

Large language models (LLMs) hold promise in clinical decision support but face major challenges in safety evaluation and effectiveness validation. We developed the Clinical Safety-Effectiveness Dual-Track Benchmark (CSEDB), a multidimensional framework built on clinical expert consensus, encompassing 30 criteria covering critical areas like critical illness recognition, guideline adherence, and medication safety, with weighted consequence measures. Thirty-two specialist physicians developed and reviewed 2,069 open-ended Q&A items aligned with these criteria, spanning 26 clinical departments to simulate real-world scenarios. Benchmark testing of six LLMs revealed moderate overall performance (average total score 57.2%, safety 54.7%, effectiveness 62.3%), with a significant 13.3% performance drop in high-risk scenarios (p < 0.0001). Domain-specific medical LLMs showed consistent performance advantages over general-purpose models, with relatively higher top scores in safety (0.912) and effectiveness (0.861). The findings of this study not only provide a standardized metric for evaluating the clinical application of medical LLMs, facilitating comparative analyses, risk exposure identification, and improvement directions across different scenarios, but also hold the potential to promote safer and more effective deployment of large language models in healthcare environments.

cs.CL↗

arXiv2025

MVAR: MultiVariate AutoRegressive Air Pollutants Forecasting Model

Air pollutants pose a significant threat to the environment and human health, thus forecasting accurate pollutant concentrations is essential for pollution warnings and policy-making. Existing studies predominantly focus on single-pollutant forecasting, neglecting the interactions among different pollutants and their diverse spatial responses. To address the practical needs of forecasting multivariate air pollutants, we propose MultiVariate AutoRegressive air pollutants forecasting model (MVAR), which reduces the dependency on long-time-window inputs and boosts the data utilization efficiency. We also design the Multivariate Autoregressive Training Paradigm, enabling MVAR to achieve 120-hour long-term sequential forecasting. Additionally, MVAR develops Meteorological Coupled Spatial Transformer block, enabling the flexible coupling of AI-based meteorological forecasts while learning the interactions among pollutants and their diverse spatial responses. As for the lack of standardized datasets in air pollutants forecasting, we construct a comprehensive dataset covering 6 major pollutants across 75 cities in North China from 2018 to 2023, including ERA5 reanalysis data and FuXi-2.0 forecast data. Experimental results demonstrate that the proposed model outperforms state-of-the-art methods and validate the effectiveness of the proposed architecture.

cs.CV↗

arXiv2025

Rainbow Artifacts from Electromagnetic Signal Injection Attacks on Image Sensors

Image sensors are integral to a wide range of safety- and security-critical systems, including surveillance infrastructure, autonomous vehicles, and industrial automation. These systems rely on the integrity of visual data to make decisions. In this work, we investigate a novel class of electromagnetic signal injection attacks that target the analog domain of image sensors, allowing adversaries to manipulate raw visual inputs without triggering conventional digital integrity checks. We uncover a previously undocumented attack phenomenon on CMOS image sensors: rainbow-like color artifacts induced in images captured by image sensors through carefully tuned electromagnetic interference. We further evaluate the impact of these attacks on state-of-the-art object detection models, showing that the injected artifacts propagate through the image signal processing pipeline and lead to significant mispredictions. Our findings highlight a critical and underexplored vulnerability in the visual perception stack, highlighting the need for more robust defenses against physical-layer attacks in such systems.

cs.CR↗