arXiv subjects
Publications and source records attributed to Yong Wang.
This paper aims to derive new anomaly cancellation formulas by combining modular forms with E8 and E8*E8 bundles. To this end, we systematically twist and generalize known SL(2,Z) modular forms to define new modular forms associated with these bundles on odd-dimensional spin and spin^c manifolds, leading to a new series of anomaly cancellation formulas.
Meta-analyses and systematic reviews demand rigorous abductive reasoning to build, test, and refine hypotheses across vast, heterogeneous literature. While NLP advancements have automated parts of this pipeline, existing tools often detach researchers from the cognitive loop or function merely as retrieval engines, leading to loss of intellectual ownership and frequent context switching. We present Research IDE, a prototype reimagining authoring environments through the "Research as Code" metaphor. Research IDE embeds a multi-agent backend into the writing flow, enabling in-situ verification via "hypothesis breakpoints." A one-week field deployment with 8 domain experts, followed by a reflective workshop, as a Research through Design (RtD) probe, reveals that users strongly preferred this verification workflow, actively leveraged prior knowledge for confirmation, and reported that breakpoints sparked insights. Drawing from participant feedback and suggestions, we derive design implications for future AI-assisted research tools that fully preserve researcher autonomy and intellectual ownership while harnessing computational scale.
Visual instruction tuning is a key training stage of large multimodal models. However, when learning multiple visual tasks simultaneously, this approach often results in suboptimal and imbalanced overall performance due to latent knowledge conflicts across tasks. To mitigate this issue, we propose a novel Adaptive Task Balancing approach tailored for visual instruction tuning (VisATB). Specifically, we measure two critical dimensions for visual task balancing based on validation performance: (1) Inter-Task Contribution, the mechanism where learning one task enhances the performance on others owing to shared knowledge across tasks, and (2) Intra-Task Difficulty, which denotes the inherent learning difficulty of a single task. Furthermore, we propose prioritizing three categories of tasks with greater weight: those that offer substantial contributions to others, those that receive minimal contributions from others, and those that present high learning difficulties. Among these three task weighting strategies, the first and third focus on improving overall performance, and the second targets the mitigation of performance imbalance. Extensive experiments on three benchmarks demonstrate that our VisATB approach consistently achieves superior and more balanced overall performance in visual instruction tuning. The data, code, and models are available at https://github.com/YanqiDai/VisATB.
Anomaly detection of multi-temporal modal data in Wireless Sensor Network (WSN) can provide an important guarantee for reliable network operation. Existing anomaly detection methods in multi-temporal modal data scenarios have the problems of insufficient extraction of spatio-temporal correlation features, high cost of anomaly sample category annotation, and imbalance of anomaly samples. In this paper, a graph neural network anomaly detection backbone network incorporating spatio-temporal correlation features and a multi-task self-supervised training strategy of "pre-training - graph prompting - fine-tuning" are designed for the characteristics of WSN graph structure data. First, the anomaly detection backbone network is designed by improving the Mamba model based on a multi-scale strategy and inter-modal fusion method, and combining it with a variational graph convolution module, which is capable of fully extracting spatio-temporal correlation features in the multi-node, multi-temporal modal scenarios of WSNs. Secondly, we design a three-subtask learning "pre-training" method with no-negative comparative learning, prediction, and reconstruction to learn generic features of WSN data samples from unlabeled data, and design a "graph prompting-fine-tuning" mechanism to guide the pre-trained self-supervised learning. The model is fine-tuned through the "graph prompting-fine-tuning" mechanism to guide the pre-trained self-supervised learning model to complete the parameter fine-tuning, thereby reducing the training cost and enhancing the detection generalization performance. The F1 metrics obtained from experiments on the public dataset and the actual collected dataset are up to 91.30% and 92.31%, respectively, which provides better detection performance and generalization ability than existing methods designed by the method.
Fake News Video Detection (FNVD) is critical for social stability. Existing methods typically assume consistent news topic distribution between training and test phases, failing to detect fake news videos tied to emerging events and unseen topics. To bridge this gap, we introduce RADAR, the first framework that enables test-time adaptation to unseen news videos. RADAR pioneers a new retrieval-guided adaptation paradigm that leverages stable (source-close) videos from the target domain to guide robust adaptation of semantically related but unstable instances. Specifically, we propose an Entropy Selection-Based Retrieval mechanism that provides videos with stable (low-entropy), relevant references for adaptation. We also introduce a Stable Anchor-Guided Alignment module that explicitly aligns unstable instances' representations to the source domain via distribution-level matching with their stable references, mitigating severe domain discrepancies. Finally, our novel Target-Domain Aware Self-Training paradigm can generate informative pseudo-labels augmented by stable references, capturing varying and imbalanced category distributions in the target domain and enabling RADAR to adapt to the fast-changing label distributions. Extensive experiments demonstrate that RADAR achieves superior performance for test-time FNVD, enabling strong on-the-fly adaptation to unseen fake news video topics.
Omni-modal models that have multimodal input and output are emerging. However, benchmarking their multimodal generation, especially in image generation, is challenging due to the subtleties of human preferences and model biases. Many image generation benchmarks focus on aesthetics instead of the fine-grained generation capabilities of these models, failing to evaluate their visual intelligence with objective metrics. In PixelArena, we propose using semantic segmentation tasks to objectively examine their fine-grained generative intelligence with pixel precision. With our benchmark and experiments, we find the latest Gemini 3 Pro Image has emergent image generation capabilities that generate semantic masks with high fidelity under zero-shot settings, showcasing visual intelligence unseen before and true generalization in new image generation tasks. We further investigate its results, compare them qualitatively and quantitatively with those of other models, and present failure cases. The findings not only signal exciting progress in the field but also provide insights into future research related to dataset development, omni-modal model development, and the design of metrics.
A kind of two-variable elliptic genus for almost-complex manifolds was introduced by Ping Li and its various properties were established by him. In this paper, we define a two-variable elliptic genus for odd dimensional spin manifolds which is the index for some Toeplitz operator and a holomorphic $SL(2,Z)$-Jacobi form. We also define some two-variable elliptic genera for almost-complex manifolds and odd dimensional spin manifolds which are holomorphic $Γ_0(2)$, $Γ^0(2)$, $Γ_θ$-Jacobi forms. By these Jacobi forms, we can get some $SL(2,{\bf Z})$ and $Γ^0(2)$ modular forms. By these $SL(2,{\bf Z})$ and $Γ^0(2)$ modular forms, we get some interesting anomaly cancellation formulas for almost complex manifolds and odd spin manifolds. As corollaries, we get some divisibility results of the holomorphic Euler characteristic number and the index of Toeplitz operators. In addition, we also define some another two-variable elliptic genera for even (rep. odd ) dimensional manifolds which are meromorphic $Γ_0(2)$, $Γ^0(2)$, $Γ_θ$-Jacobi forms.
The image geolocalization task aims to predict the location where an image was taken anywhere on Earth using visual clues. Existing large vision-language model (LVLM) approaches leverage world knowledge, chain-of-thought reasoning, and agentic capabilities, but overlook a common strategy used by humans -- using maps. In this work, we first equip the model \textit{Thinking with Map} ability and formulate it as an agent-in-the-map loop. We develop a two-stage optimization scheme for it, including agentic reinforcement learning (RL) followed by parallel test-time scaling (TTS). The RL strengthens the agentic capability of model to improve sampling efficiency, and the parallel TTS enables the model to explore multiple candidate paths before making the final prediction, which is crucial for geolocalization. To evaluate our method on up-to-date and in-the-wild images, we further present MAPBench, a comprehensive geolocalization training and evaluation benchmark composed entirely of real-world images. Experimental results show that our method outperforms existing open- and closed-source models on most metrics, specifically improving Acc@500m from 8.0\% to 22.1\% compared to \textit{Gemini-3-Pro} with Google Search/Map grounded mode.
Dark mode has gained widespread adoption across mobile platforms due to its benefits in reducing eye strain and conserving battery life. However, while the mobile system switches to dark mode, most visualizations remain designed for light mode, causing visual disruptions. Existing methods, such as manual adjustment or color inversion, are either time-consuming or fail to preserve the semantic meaning of colors in visualizations, making them less effective in dark mode. To address this challenge, we propose Chameleon, an algorithm that automatically transforms light mode visualizations into dark mode while maintaining visual clarity and color semantics. By optimizing for luminance contrast, color consistency, and adjacent color differences, Chameleon ensures that the transformed visualizations are legible and visually coherent. Our evaluation includes case study, expert interview, system evaluation, and a user study, and these demonstrate that Chameleon is effective at translating visualizations for dark mode.
Motivated by the work of D. Hoff and K. Zumbrun (Indiana Univ. Math. J. 44: 603-676, 1995), we investigate the diffusion wave phenomena in three-dimensional incompressible viscoelastic flows. By employing the representation formula of the wave equation and the stationary phase methods on the sphere $\mathbb{S}^{d-1}$, we establish $L^p$ decay estimates for the solution over the whole range $1\leq p \leq \infty$, which reveals the hyperbolic nature of the incompressible viscoelastic flows.
Multimodal Large Language Model (MLLM) Personalization is a critical research problem that facilitates personalized dialogues with MLLMs targeting specific entities (known as personalized concepts). However, existing methods and benchmarks focus on the simple, context-agnostic visual identification and textual replacement of the personalized concept (e.g., "A yellow puppy" -> "Your puppy Mochi"), overlooking the ability to support long-context conversations. An ideal personalized MLLM assistant is capable of engaging in long-context dialogues with humans and continually improving its experience quality by learning from past dialogue histories. To bridge this gap, we propose LCMP, the first Long-Context MLLM Personalization evaluation benchmark. LCMP assesses the capability of MLLMs in perceiving variations of personalized concepts and generating contextually appropriate personalized responses that reflect these variations. As a strong baseline for LCMP, we introduce a novel training-free and state-aware framework TAME. TAME endows MLLMs with double memories to manage the temporal and persistent variations of each personalized concept in a differentiated manner. In addition, TAME incorporates a new training-free Retrieve-then-Align Augmented Generation (RA2G) paradigm. RA2G introduces an alignment step to extract the contextually fitted information from the multi-memory retrieved knowledge to the current questions, enabling better interactions for complex real-world user queries. Experiments on LCMP demonstrate that TAME achieves the best performance, showcasing remarkable and evolving interaction experiences in long-context scenarios.
In this paper, we investigate the existence of 2-D Taylor-Couette flow for a rarefied gas between two coaxial rotating cylinders, characterized by differing angular velocities at the outer boundary $\{r=1\}$ and the inner boundary $\{r=r_{1}>0\}$, with a small relative strength denoted by $α$. We formulate the problem using the steady Boltzmann equation in polar coordinates and seek a solution invariant under rotation. We assume that the steady state has the specific form $F(r,v_{r},v_ϕ-α\frac{r-r_{1}}{1-r_{1}},v_{z})$, where the translation angular velocity $α\frac{r-r_{1}}{1-r_{1}}$ is linearly sheared along the radial direction. With this ansatz, the problem is reduced to solve the nonlinear steady Boltzmann equation with geometric correction, subject to an external shear force of strength $α$ and the homogeneous non-moving diffuse reflection boundary condition. We establish the existence of a non-equilibrium steady solution for any small enough shear rate $α$ through Caflisch's decomposition, complemented by careful uniform estimates based on Guo's $L^{\infty} \cap L^2$ framework. The steady profile displays a polynomial tail behavior at large velocities. For the proof, we develop a delicate double-parameter $(ε,σ)$-approximation argument for the construction of solutions. In particular, we obtain uniform macroscopic dissipation estimates in the absence of mass conservation for $σ\in [0,1)$ getting close to 1. Additionally, due to the non-trivial geometric effects, we develop subtle constructions of test functions by solving second-order ODEs with geometric corrections to establish macroscopic dissipation. Furthermore, we justify the non-negativity of the steady profile by demonstrating its large-time asymptotic stability with an exponential convergence rate under radial perturbations.
Using Liu's modular invariance method and its odd-dimensional extension by Han and Yu, we establish new Witten rigidity theorems for the generalized Witten genus of twisted Dirac operators on even-dimensional spin$^c$ manifolds and twisted Toeplitz operators on odd-dimensional spin$^c$ manifolds with circle actions.
Quantum Neural Networks (QNNs) represent a promising fusion of quantum computing and neural network architectures, offering speed-ups and efficient processing of high-dimensional, entangled data. A crucial component of QNNs is the encoder, which maps classical input data into quantum states. However, choosing suitable encoders remains a significant challenge, largely due to the lack of systematic guidance and the trial-and-error nature of current approaches. This process is further impeded by two key challenges: (1) the difficulty in evaluating encoded quantum states prior to training, and (2) the lack of intuitive methods for analyzing an encoder's ability to effectively distinguish data features. To address these issues, we introduce a novel visualization tool, XQAI-Eyes, which enables QNN developers to compare classical data features with their corresponding encoded quantum states and to examine the mixed quantum states across different classes. By bridging classical and quantum perspectives, XQAI-Eyes facilitates a deeper understanding of how encoders influence QNN performance. Evaluations across diverse datasets and encoder designs demonstrate XQAI-Eyes's potential to support the exploration of the relationship between encoder design and QNN effectiveness, offering a holistic and transparent approach to optimizing quantum encoders. Moreover, domain experts used XQAI-Eyes to derive two key practices for quantum encoder selection, grounded in the principles of pattern preservation and feature mapping.
Video diffusion models (VDMs) perform attention computation over the 3D spatio-temporal domain. Compared to large language models (LLMs) processing 1D sequences, their memory consumption scales cubically, necessitating parallel serving across multiple GPUs. Traditional parallelism strategies partition the computational graph, requiring frequent high-dimensional activation transfers that create severe communication bottlenecks. To tackle this issue, we exploit the local spatio-temporal dependencies inherent in the diffusion denoising process and propose Latent Parallelism (LP), the first parallelism strategy tailored for VDM serving. \textcolor{black}{LP decomposes the global denoising problem into parallelizable sub-problems by dynamically rotating the partitioning dimensions (temporal, height, and width) within the compact latent space across diffusion timesteps, substantially reducing the communication overhead compared to prevailing parallelism strategies.} To ensure generation quality, we design a patch-aligned overlapping partition strategy that matches partition boundaries with visual patches and a position-aware latent reconstruction mechanism for smooth stitching. Experiments on three benchmarks demonstrate that LP reduces communication overhead by up to 97\% over baseline methods while maintaining comparable generation quality. As a non-intrusive plug-in paradigm, LP can be seamlessly integrated with existing parallelism strategies, enabling efficient and scalable video generation services.
Unmanned Aerial Vehicles (UAVs) offer wide-ranging applications but also pose significant safety and privacy violation risks in areas like airport and infrastructure inspection, spurring the rapid development of Anti-UAV technologies in recent years. However, current Anti-UAV research primarily focuses on RGB, infrared (IR), or RGB-IR videos captured by fixed ground cameras, with little attention to tracking target UAVs from another moving UAV platform. To fill this gap, we propose a new multi-modal visual tracking task termed UAV-Anti-UAV, which involves a pursuer UAV tracking a target adversarial UAV in the video stream. Compared to existing Anti-UAV tasks, UAV-Anti-UAV is more challenging due to severe dual-dynamic disturbances caused by the rapid motion of both the capturing platform and the target. To advance research in this domain, we construct a million-scale dataset consisting of 1,810 videos, each manually annotated with bounding boxes, a language prompt, and 15 tracking attributes. Furthermore, we propose MambaSTS, a Mamba-based baseline method for UAV-Anti-UAV tracking, which enables integrated spatial-temporal-semantic learning. Specifically, we employ Mamba and Transformer models to learn global semantic and spatial features, respectively, and leverage the state space model's strength in long-sequence modeling to establish video-level long-term context via a temporal token propagation mechanism. We conduct experiments on the UAV-Anti-UAV dataset to validate the effectiveness of our method. A thorough experimental evaluation of 50 modern deep tracking algorithms demonstrates that there is still significant room for improvement in the UAV-Anti-UAV domain. The dataset and codes will be available at {\color{magenta}https://github.com/983632847/Awesome-Multimodal-Object-Tracking}.
In-depth analysis of competitive debates is essential for participants to develop argumentative skills and refine strategies, and further improve their debating performance. However, manual analysis of unstructured and unlabeled textual records of debating is time-consuming and ineffective, as it is challenging to reconstruct contextual semantics and track logical connections from raw data. To address this, we propose Conch, an interactive visualization system that systematically analyzes both what is debated and how it is debated. In particular, we propose a novel parallel spiral visualization that compactly traces the multidimensional evolution of clash points and participant interactions throughout debate process. In addition, we leverage large language models with well-designed prompts to automatically identify critical debate elements such as clash points, disagreements, viewpoints, and strategies, enabling participants to understand the debate context comprehensively. Finally, through two case studies on real-world debates and a carefully-designed user study, we demonstrate Conch's effectiveness and usability for competitive debate analysis.
Photons are promising candidates for quantum information technology due to their high robustness and long coherence time at room temperature. Inspired by the prosperous development of photonic computing techniques, recent research has turned attention to performing quantum machine learning on photonic platforms. Although photons possess a high-dimensional quantum feature space suitable for computation, a general understanding of how to harness it for learning tasks remains blank. Here, we establish both theoretically and experimentally a scalable advantage in quantum machine learning with multi-photon states. Firstly, we prove that the learning capacity of linear optical circuits scales polynomially with the photon number, enabling generalization from smaller training datasets and yielding lower test loss values. Moreover, we experimentally corroborate these findings through unitary learning and metric learning tasks, by performing online training on a fully programmable photonic integrated platform. Our work highlights the potential of photonic quantum machine learning and paves the way for achieving quantum enhancement in practical machine learning applications.