arXiv subjects

Yuxuan Zhou

Publications and source records attributed to Yuxuan Zhou.

At least 73 records · Page 4Linked to original sources

arXiv2025

JPRO: Automated Multimodal Jailbreaking via Multi-Agent Collaboration Framework

The widespread application of large VLMs makes ensuring their secure deployment critical. While recent studies have demonstrated jailbreak attacks on VLMs, existing approaches are limited: they require either white-box access, restricting practicality, or rely on manually crafted patterns, leading to poor sample diversity and scalability. To address these gaps, we propose JPRO, a novel multi-agent collaborative framework designed for automated VLM jailbreaking. It effectively overcomes the shortcomings of prior methods in attack diversity and scalability. Through the coordinated action of four specialized agents and its two core modules: Tactic-Driven Seed Generation and Adaptive Optimization Loop, JPRO generates effective and diverse attack samples. Experimental results show that JPRO achieves over a 60\% attack success rate on multiple advanced VLMs, including GPT-4o, significantly outperforming existing methods. As a black-box attack approach, JPRO not only uncovers critical security vulnerabilities in multimodal models but also offers valuable insights for evaluating and enhancing VLM robustness.

cs.CR↗

arXiv2025

LongWeave: A Long-Form Generation Benchmark Bridging Real-World Relevance and Verifiability

Generating long, informative, and factual outputs remains a major challenge for Large Language Models (LLMs). Existing benchmarks for long-form generation typically assess real-world queries with hard-to-verify metrics or use synthetic setups that ease evaluation but overlook real-world intricacies. In this paper, we introduce \textbf{LongWeave}, which balances real-world and verifiable assessment with Constraint-Verifier Evaluation (CoV-Eval). CoV-Eval constructs tasks by first defining verifiable targets within real-world scenarios, then systematically generating corresponding queries, textual materials, and constraints based on these targets. This ensures that tasks are both realistic and objectively assessable, enabling rigorous assessment of model capabilities in meeting complex real-world constraints. LongWeave supports customizable input/output lengths (up to 64K/8K tokens) across seven distinct tasks. Evaluation on 23 LLMs shows that even state-of-the-art models encounter significant challenges in long-form generation as real-world complexity and output length increase.

cs.CL↗

arXiv2025

Uni-MuMER: Unified Multi-Task Fine-Tuning of Vision-Language Model for Handwritten Mathematical Expression Recognition

Handwritten Mathematical Expression Recognition (HMER) remains a persistent challenge in Optical Character Recognition (OCR) due to the inherent freedom of symbol layouts and variability in handwriting styles. Prior methods have faced performance bottlenecks by proposing isolated architectural modifications, making them difficult to integrate coherently into a unified framework. Meanwhile, recent advances in pretrained vision-language models (VLMs) have demonstrated strong cross-task generalization, offering a promising foundation for developing unified solutions. In this paper, we introduce Uni-MuMER, which fully fine-tunes a VLM for the HMER task without modifying its architecture, effectively injecting domain-specific knowledge into a generalist framework. Our method integrates three data-driven tasks: Tree-Aware Chain-of-Thought (Tree-CoT) for structured spatial reasoning, Error-Driven Learning (EDL) for reducing confusion among visually similar characters, and Symbol Counting (SC) for improving recognition consistency in long expressions. Experiments on the CROHME and HME100K datasets show that Uni-MuMER achieves super state-of-the-art performance, outperforming the best lightweight specialized model SSAN by 16.31\% and the top-performing VLM Gemini2.5-flash by 24.42\% under zero-shot setting. Our datasets, models, and code are open-sourced at: {https://github.com/BFlameSwift/Uni-MuMER

cs.CV↗

arXiv2025

BrowserAgent: Building Web Agents with Human-Inspired Web Browsing Actions

Efficiently solving real-world problems with LLMs increasingly hinges on their ability to interact with dynamic web environments and autonomously acquire external information. While recent research like Search-R1 and WebDancer demonstrates strong performance in solving web tasks, they heavily rely on additional tools to convert the interactive web environment into static text content. This is in contrast to human browsing behaviors, which involve diverse interactions with the browser, such as scrolling, clicking, and typing. In this paper, we propose BrowserAgent, a more interactive agent that solves complex tasks through human-inspired browser actions. BrowserAgent operates directly on raw web pages via Playwright through a set of predefined browser actions. We adopt a two-stage training (Supervised Fine-Tuning (SFT) and Rejection Fine-Tuning (RFT)) to improve the model's generalization abilities. Despite using significantly less training data than Search-R1, BrowserAgent achieves more competitive results across different Open-QA tasks. Additionally, we introduce an explicit memory mechanism to store key conclusions across steps, further enhancing the model's reasoning capabilities for long-horizon tasks. Notably, BrowserAgent-7B can achieve around 20\% improvement over Search-R1 on multi-hop QA tasks like HotpotQA, 2Wiki, and Bamboogle. These results indicate that BrowserAgent can serve as a more advanced framework for more interactive and scalable web agents.

cs.CL↗

arXiv2025

Dephasing-assisted diffusive dynamics in superconducting quantum circuits

Random fluctuations caused by environmental noise can lead to decoherence in quantum systems. Exploring and controlling such dissipative processes is both fundamentally intriguing and essential for harnessing quantum systems to gain practical advantages and deeper insights. In this work, we first demonstrate the diffusive dynamics assisted by controlled dephasing noise in superconducting quantum circuits, contrasting with coherent evolution. We show that dephasing can give distinct dynamical behavior in a superconducting qubit array with quasiperiodic order. Furthermore, by preparing different excitation distributions in the qubit array, we observe that a more localized initial state relaxes to a uniformly distributed mixed state faster with dephasing noise, illustrating another counterintuitive phenomenon called Mpemba-effect-like quantum dynamics, i.e., a far-from-equilibrium state can relax toward the equilibrium faster. These results deepen our understanding of diffusive dynamics at the microscopic level, and demonstrate controlled dissipative processes as a valuable tool for investigating Markovian open quantum systems.

quant-ph↗

arXiv2025

High-performance multiplexed readout of superconducting qubits with a tunable broadband Purcell filter

Fast, high-fidelity, and low back-action readout plays a crucial role in the advancement of quantum error correction (QEC). Here, we demonstrate high-performance multiplexed readout of superconducting qubits using a tunable broadband Purcell filter, effectively resolving the fundamental trade-off between measurement speed and photon-noise-induced dephasing. By dynamically tuning the filter parameters, we suppress photon-noise-induced dephasing by a factor of 7 in idle status, while enabling rapid, high-fidelity readout in measurement status. We achieve 99.6\% single-shot readout fidelity with 100~ns readout pulse, limited primarily by relaxation errors during readout. Using a multilevel readout protocol, we further attain 99.9\% fidelity in 50~ns. Simultaneous readout of three qubits using 100~ns pulses achieves an average fidelity of 99.5\% with low crosstalk. Additionally, the readout exhibits high quantum-nondemolition (QND) performance: 99.4\% fidelity over repeated measurements and a low leakage rate below 0.1\%. Building on the tunable broadband filter, we further propose a scalable readout scheme for surface code QEC with enhanced multiplexing capability, offering a promising solution for fast and scalable QEC.

quant-ph↗

arXiv2025

Synthetic $π$-flux system in 2D superconducting qubit array with tunable coupling

Flat-band systems provide an ideal platform for exploring exotic quantum phenomena, where the strongly suppressed kinetic energy in these flat energy bands suggests the potential for exotic phases driven by geometric structure, disorder, and interactions. While intriguing phenomena and physical mechanisms have been unveiled in theoretical models, synthesizing such systems within scalable quantum platforms remains challenging. Here, we present the experimental realization of a $π$-flux rhombic system using a two-dimensional superconducting qubit array with tunable coupling. We experimentally observe characteristic dynamics, e.g., $π$-flux driven destructive interference, and demonstrate the protocol for eigenstate preparation in this rhombic array with coupler-assisted flux. Our results provide future possibilities for exploring the interplay of geometry, interactions, and quantum information encoding in such degenerate systems.

quant-ph↗

arXiv2025

State Similarity in Modular Superconducting Quantum Processors with Classical Communications

As quantum devices continue to scale, distributed quantum computing emerges as a promising strategy for executing large-scale tasks across modular quantum processors. A central challenge in this paradigm is verifying the correctness of computational outcomes when subcircuits are executed independently following circuit cutting. Here we propose a cross-platform fidelity estimation algorithm tailored for modular architectures. Our method achieves substantial reductions in sample complexity compared to previous approaches designed for single-processor systems. We experimentally implement the protocol on modular superconducting quantum processors with up to 6 qubits to verify the similarity of two 11-qubit GHZ states. Beyond verification, we show that our algorithm enables a federated quantum kernel method that preserves data privacy. As a proof of concept, we apply it to a 5-qubit quantum phase learning task using six 3-qubit modules, successfully extracting phase information with just eight training samples. These results establish a practical path for scalable verification and trustworthy quantum machine learning of modular quantum processors.

quant-ph↗

arXiv2025

Evaluating LLMs Across Multi-Cognitive Levels: From Medical Knowledge Mastery to Scenario-Based Problem Solving

Large language models (LLMs) have demonstrated remarkable performance on various medical benchmarks, but their capabilities across different cognitive levels remain underexplored. Inspired by Bloom's Taxonomy, we propose a multi-cognitive-level evaluation framework for assessing LLMs in the medical domain in this study. The framework integrates existing medical datasets and introduces tasks targeting three cognitive levels: preliminary knowledge grasp, comprehensive knowledge application, and scenario-based problem solving. Using this framework, we systematically evaluate state-of-the-art general and medical LLMs from six prominent families: Llama, Qwen, Gemma, Phi, GPT, and DeepSeek. Our findings reveal a significant performance decline as cognitive complexity increases across evaluated models, with model size playing a more critical role in performance at higher cognitive levels. Our study highlights the need to enhance LLMs' medical capabilities at higher cognitive levels and provides insights for developing LLMs suited to real-world medical applications.

cs.CL↗

arXiv2025

An evaluation of LLMs for generating movie reviews: GPT-4o, Gemini-2.0 and DeepSeek-V3

Large language models (LLMs) have been prominent in various tasks, including text generation and summarisation. The applicability of LLMs to the generation of product reviews is gaining momentum, paving the way for the generation of movie reviews. In this study, we propose a framework that generates movie reviews using three LLMs (GPT-4o, DeepSeek-V3, and Gemini-2.0), and evaluate their performance by comparing the generated outputs with IMDb user reviews. We use movie subtitles and screenplays as input to the LLMs and investigate how they affect the quality of reviews generated. We review the LLM-based movie reviews in terms of vocabulary, sentiment polarity, similarity, and thematic consistency in comparison to IMDB user reviews. The results demonstrate that LLMs are capable of generating syntactically fluent and structurally complete movie reviews. Nevertheless, there is still a noticeable gap in emotional richness and stylistic coherence between LLM-generated and IMDb reviews, suggesting that further refinement is needed to improve the overall quality of movie review generation. We provided a survey-based analysis where participants were told to distinguish between LLM and IMDb user reviews. The results show that LLM-generated reviews are difficult to distinguish from IMDB user reviews. We found that DeepSeek-V3 produced the most balanced reviews, closely matching IMDb reviews. GPT-4o overemphasised positive emotions, while Gemini-2.0 captured negative emotions better but showed excessive emotional intensity.

cs.CL↗

arXiv2025

Predicting Diabetic Macular Edema Treatment Responses Using OCT: Dataset and Methods of APTOS Competition

Diabetic macular edema (DME) significantly contributes to visual impairment in diabetic patients. Treatment responses to intravitreal therapies vary, highlighting the need for patient stratification to predict therapeutic benefits and enable personalized strategies. To our knowledge, this study is the first to explore pre-treatment stratification for predicting DME treatment responses. To advance this research, we organized the 2nd Asia-Pacific Tele-Ophthalmology Society (APTOS) Big Data Competition in 2021. The competition focused on improving predictive accuracy for anti-VEGF therapy responses using ophthalmic OCT images. We provided a dataset containing tens of thousands of OCT images from 2,000 patients with labels across four sub-tasks. This paper details the competition's structure, dataset, leading methods, and evaluation metrics. The competition attracted strong scientific community participation, with 170 teams initially registering and 41 reaching the final round. The top-performing team achieved an AUC of 80.06%, highlighting the potential of AI in personalized DME treatment and clinical decision-making.

eess.IV↗

arXiv2025

S3MOT: Monocular 3D Object Tracking with Selective State Space Model

Accurate and reliable multi-object tracking (MOT) in 3D space is essential for advancing robotics and computer vision applications. However, it remains a significant challenge in monocular setups due to the difficulty of mining 3D spatiotemporal associations from 2D video streams. In this work, we present three innovative techniques to enhance the fusion and exploitation of heterogeneous cues for monocular 3D MOT: (1) we introduce the Hungarian State Space Model (HSSM), a novel data association mechanism that compresses contextual tracking cues across multiple paths, enabling efficient and comprehensive assignment decisions with linear complexity. HSSM features a global receptive field and dynamic weights, in contrast to traditional linear assignment algorithms that rely on hand-crafted association costs. (2) We propose Fully Convolutional One-stage Embedding (FCOE), which eliminates ROI pooling by directly using dense feature maps for contrastive learning, thus improving object re-identification accuracy under challenging conditions such as varying viewpoints and lighting. (3) We enhance 6-DoF pose estimation through VeloSSM, an encoder-decoder architecture that models temporal dependencies in velocity to capture motion dynamics, overcoming the limitations of frame-based 3D inference. Experiments on the KITTI public test benchmark demonstrate the effectiveness of our method, achieving a new state-of-the-art performance of 76.86~HOTA at 31~FPS. Our approach outperforms the previous best by significant margins of +2.63~HOTA and +3.62~AssA, showcasing its robustness and efficiency for monocular 3D MOT tasks. The code and models are available at https://github.com/bytepioneerX/s3mot.

cs.CV↗

arXiv2025

Secure Transfer Learning: Training Clean Models Against Backdoor in (Both) Pre-trained Encoders and Downstream Datasets

Transfer learning from pre-trained encoders has become essential in modern machine learning, enabling efficient model adaptation across diverse tasks. However, this combination of pre-training and downstream adaptation creates an expanded attack surface, exposing models to sophisticated backdoor embeddings at both the encoder and dataset levels--an area often overlooked in prior research. Additionally, the limited computational resources typically available to users of pre-trained encoders constrain the effectiveness of generic backdoor defenses compared to end-to-end training from scratch. In this work, we investigate how to mitigate potential backdoor risks in resource-constrained transfer learning scenarios. Specifically, we conduct an exhaustive analysis of existing defense strategies, revealing that many follow a reactive workflow based on assumptions that do not scale to unknown threats, novel attack types, or different training paradigms. In response, we introduce a proactive mindset focused on identifying clean elements and propose the Trusted Core (T-Core) Bootstrapping framework, which emphasizes the importance of pinpointing trustworthy data and neurons to enhance model security. Our empirical evaluations demonstrate the effectiveness and superiority of T-Core, specifically assessing 5 encoder poisoning attacks, 7 dataset poisoning attacks, and 14 baseline defenses across five benchmark datasets, addressing four scenarios of 3 potential backdoor threats.

cs.LG↗

arXiv2025

Logical multi-qubit entanglement with dual-rail superconducting qubits

Recent advances in quantum error correction (QEC) across hardware platforms have demonstrated operation near and beyond the fault-tolerance threshold, yet achieving exponential suppression of logical errors through code scaling remains a critical challenge. Erasure qubits, which enable hardware-level detection of dominant error types, offer a promising path toward resource-efficient QEC by exploiting error bias. Single erasure qubits with dual-rail encoding in superconducting cavities and transmons have demonstrated high coherence and low single-qubit gate errors with mid-circuit erasure detection, but the generation of multi-qubit entanglement--a fundamental requirement for quantum computation and error correction--has remained an outstanding milestone. Here, we demonstrate a superconducting processor integrating four dual-rail erasure qubits that achieves the logical multi-qubit entanglement with error-biased protection. Each dual-rail qubit, encoded in pairs of tunable transmons, preserves millisecond-scale coherence times and single-qubit gate errors at the level of $10^{-5}$. By engineering tunable couplings between logical qubits, we generate high-fidelity entangled states resilient to physical qubit noise, including logical Bell states (98.8% fidelity) and a three-logical-qubit Greenberger-Horne-Zeilinger (GHZ) state (93.5% fidelity). A universal gate set is realized through a calibrated logical controlled-NOT (CNOT) gate with 96.2% process fidelity, enabled by coupler-activated $XX$ interactions in the protected logical subspace. This work advances dual-rail architectures beyond single-qubit demonstrations, providing a blueprint for concatenated quantum error correction with erasure qubits.

quant-ph↗

arXiv2025

Contrasting magnetism in VPS3 and CrI3 monolayers with the common honeycomb S = 3/2 spin lattice

Two-dimensional (2D) magnetic materials are promising candidates for spintronics and quantum technologies. One extensively studied example is the ferromagnetic (FM) CrI$_3$ monolayer with the honeycomb Cr$^{3+}$ ($t_{2g}^3$, $S$ = 3/2) spin lattice, while VPS$_3$ has a same honeycomb $S$ = 3/2 spin lattice (V$^{2+}$, $t_{2g}^3$) but displays N$\acute{e}$el antiferromagnetism (AFM). In this work, we study the electronic structure and particularly the contrasting magnetism of VPS$_3$ and CrI$_3$ monolayers. We find that VPS$_3$ is a Mott-Hubbard insulator but CrI$_3$ is a charge-transfer insulator, and therefore their magnetic exchange mechanisms are essentially different. The first nearest-neighbor (1NN) direct $d$-$d$ exchange dominates in VPS$_3$, thus leading to a strong antiferromagnetic (AF) coupling. However, the formation of vanadium vacancies, associated with instability of the low-valence V$^{2+}$ ions, suppresses the AF coupling and thus strongly reduces the N$\acute{e}$el temperature ($T_{\text{N}}$) in line with the experimental observation. In contrast, our results reveal that the major 1NN $d$-$p$-$d$ superexchanges in CrI$_3$ via different channels give rise to competing FM and AF couplings, ultimately resulting in a weak FM coupling as observed experimentally. After revisiting several important superexchange channels reported in the literature, based on our MLWFs and tight-binding analyses, we note that some antiphase contributions must be subtly and simultaneously considered, and thus we provide a deeper insight into the FM coupling of CrI$_3$. Moreover, we identify and compare the major contributions to the magnetic anisotropy, i.e., a weak shape anisotropy in VPS$_3$ and a relatively strong exchange anisotropy in CrI$_3$.

cond-mat.mtrl-sci↗

arXiv2025

A thermal-noise-resilient microwave quantum network traversing 4 K

Quantum communication at microwave frequencies has been fundamentally constrained by the susceptibility of microwave photons to thermal noise, hindering their application in scalable quantum networks. Here we demonstrate a thermal-noise-resilient microwave quantum network that establishes coherent coupling between two superconducting qubits through a 4 K thermalized niobium-titanium transmission line. By overcoupling the communication channel to a cold load at 10 mK, we suppress the effective thermal occupancy of the channel to 0.06 photons through radiative cooling -- a two-order-of-magnitude reduction below ambient thermal noise. We then decouple the cold load and rapidly transfer microwave quantum states through the channel while it rethermalizes, achieving a 58.5% state transfer fidelity and a 52.3% Bell entanglement fidelity, both exceeding the classical communication threshold. Our architecture overcomes the temperature-compatibility barrier for microwave quantum systems, providing a scalable framework for distributed quantum computing and enabling hybrid quantum networks with higher-temperature semiconductor or photonic platforms.

quant-ph↗

arXiv2025

Multiferroic Metallic Monolayer Cu(CrSe2)2

The two-dimensional (2D) Cu(CrSe$_2$)$_2$ monolayer stands out for its combined ferromagnetic (FM), ferroelectric (FE), and metallic properties, marking itself as a prominent 2D multiferroic metal. This work studies those properties and the relevant physics, using density functional calculations, Monte Carlo simulations, and $ab$ $initio$ molecular dynamics. Our results show that Cu(CrSe$_2$)$_2$ monolayer is in the Cr$^{3+}$ $t_{2g}^3$ state with $S$ = 3/2 and Cu$^{1+}$ $3d^{10}$ with $S$ = 0. The observed in-plane magnetic anisotropy primarily arises from exchange anisotropy, which is associated with the Cr-Se-Cr itinerant ferromagnetism. In contrast, both single-ion anisotropy and shape magnetic anisotropy contribute negligibly. The Dzyaloshinskii-Moriya interaction is also quite weak, only about 3\% of the intralayer exchange parameters. Our Monte Carlo simulations show a FM Curie temperature ($T_{\rm C}$) of 190 K. Moreover, the monolayer exhibits a vertical FE polarization of 1.79 pC/m and a FE polarization switching barrier of 182 meV/f.u., and the FE state remains stable above room temperature as shown by $ab$ $initio$ molecular dynamics simulations. Furthermore, a magnetoelectric coupling is partially manifested by a magnetization rotation from in-plane to out-of-plane associated with a FE-to-paraelectric transition. The magnetization rotation can also be induced by either hole or electron doping, and the hole doping increases the $T_{\rm C}$ up to 238 K. In addition, tensile strain reduces the FE polarization but enhances $T_{\rm C}$ to 290 K, while a compressive strain gives an opposite effect. Therefore, the multiferroic metallic Cu(CrSe$_2$)$_2$ monolayer may be explored for advanced multifunctional electronic devices.

cond-mat.mtrl-sci↗

arXiv2025

Experimental Modeling of Chiral Active Robots and a Minimal Model of Non-Gaussian Displacements

We design 3D-printed motor-driven active particles and find that their dynamics can be characterized using the model of overdamped chiral active Brownian particles (ABPs), as demonstrated by measured angular statistics and translational mean squared displacements (MSDs). Furthermore, we propose a minimal model that reproduces the double-peak velocity distributions and further predicts a transition from the single-peak to the double-peak displacement distributions in short-time regimes. The model provides a clear physics picture of these phenomena, originating from the competition between the active motion and the translational diffusion. Our experiments confirm such picture. The minimal model enhances our understanding of activity-driven non-Gaussian phenomena. The designed particles could be further applied in the study of collective chiral motions.

cond-mat.soft↗