arXiv subjects

Yu Wu

Publications and source records attributed to Yu Wu.

At least 109 records · Page 6Linked to original sources

arXiv2025

CoMM: A Coherent Interleaved Image-Text Dataset for Multimodal Understanding and Generation

Interleaved image-text generation has emerged as a crucial multimodal task, aiming at creating sequences of interleaved visual and textual content given a query. Despite notable advancements in recent multimodal large language models (MLLMs), generating integrated image-text sequences that exhibit narrative coherence and entity and style consistency remains challenging due to poor training data quality. To address this gap, we introduce CoMM, a high-quality Coherent interleaved image-text MultiModal dataset designed to enhance the coherence, consistency, and alignment of generated multimodal content. Initially, CoMM harnesses raw data from diverse sources, focusing on instructional content and visual storytelling, establishing a foundation for coherent and consistent content. To further refine the data quality, we devise a multi-perspective filter strategy that leverages advanced pre-trained models to ensure the development of sentences, consistency of inserted images, and semantic alignment between them. Various quality evaluation metrics are designed to prove the high quality of the filtered dataset. Meanwhile, extensive few-shot experiments on various downstream tasks demonstrate CoMM's effectiveness in significantly enhancing the in-context learning capabilities of MLLMs. Moreover, we propose four new tasks to evaluate MLLMs' interleaved generation abilities, supported by a comprehensive evaluation framework. We believe CoMM opens a new avenue for advanced MLLMs with superior multimodal in-context learning and understanding ability.

cs.CV↗

arXiv2025

Decoupled anisotropic Charge-Phonon Transport Enables Exceptional n-Type Thermoelectric Performance in CuBiSCl$_2$

First-principles calculations demonstrate an exceptional decoupling of charge and thermal transport along the \textit{a}-axis in CuBiSCl$_2$. The material achieves superior electron mobility (138 cm$^2$/V$\cdot$s at 300 K) through delocalized Bi-6\textit{p}/S-3\textit{p} networks while maintaining ultralow lattice thermal conductivity (0.40 W/mK at 300 K) via Cu-dominated anharmonic phonon scattering - both optimized along the same crystallographic direction. This simultaneous optimization originates from the anisotropic bonding hierarchy where [BiSCl$_2$]$_n$ ribbons enable efficient charge transport along \textit{a}-axis, while the soft vibrational modes associated with Cu atoms strongly scatter heat-carrying phonons. The resulting high power factor (1.71 mW/mK$^2$ at 700 K) and peak \textit{ZT} of 1.57 establish CuBiSCl$_2$ as a model system that realizes the long-sought "phonon glass-electron crystal" paradigm through crystallographically engineered transport channels.

cond-mat.mtrl-sci↗

arXiv2025

Implicit Bias Injection Attacks against Text-to-Image Diffusion Models

The proliferation of text-to-image diffusion models (T2I DMs) has led to an increased presence of AI-generated images in daily life. However, biased T2I models can generate content with specific tendencies, potentially influencing people's perceptions. Intentional exploitation of these biases risks conveying misleading information to the public. Current research on bias primarily addresses explicit biases with recognizable visual patterns, such as skin color and gender. This paper introduces a novel form of implicit bias that lacks explicit visual features but can manifest in diverse ways across various semantic contexts. This subtle and versatile nature makes this bias challenging to detect, easy to propagate, and adaptable to a wide range of scenarios. We further propose an implicit bias injection attack framework (IBI-Attacks) against T2I diffusion models by precomputing a general bias direction in the prompt embedding space and adaptively adjusting it based on different inputs. Our attack module can be seamlessly integrated into pre-trained diffusion models in a plug-and-play manner without direct manipulation of user input or model retraining. Extensive experiments validate the effectiveness of our scheme in introducing bias through subtle and diverse modifications while preserving the original semantics. The strong concealment and transferability of our attack across various scenarios further underscore the significance of our approach. Code is available at https://github.com/Hannah1102/IBI-attacks.

cs.CV↗

arXiv2025

Re-HOLD: Video Hand Object Interaction Reenactment via adaptive Layout-instructed Diffusion Model

Current digital human studies focusing on lip-syncing and body movement are no longer sufficient to meet the growing industrial demand, while human video generation techniques that support interacting with real-world environments (e.g., objects) have not been well investigated. Despite human hand synthesis already being an intricate problem, generating objects in contact with hands and their interactions presents an even more challenging task, especially when the objects exhibit obvious variations in size and shape. To tackle these issues, we present a novel video Reenactment framework focusing on Human-Object Interaction (HOI) via an adaptive Layout-instructed Diffusion model (Re-HOLD). Our key insight is to employ specialized layout representation for hands and objects, respectively. Such representations enable effective disentanglement of hand modeling and object adaptation to diverse motion sequences. To further improve the generation quality of HOI, we design an interactive textural enhancement module for both hands and objects by introducing two independent memory banks. We also propose a layout adjustment strategy for the cross-object reenactment scenario to adaptively adjust unreasonable layouts caused by diverse object sizes during inference. Comprehensive qualitative and quantitative evaluations demonstrate that our proposed framework significantly outperforms existing methods. Project page: https://fyycs.github.io/Re-HOLD.

cs.CV↗

arXiv2025

D$^3$: Scaling Up Deepfake Detection by Learning from Discrepancy

The boom of Generative AI brings opportunities entangled with risks and concerns. Existing literature emphasizes the generalization capability of deepfake detection on unseen generators, significantly promoting the detector's ability to identify more universal artifacts. This work seeks a step toward a universal deepfake detection system with better generalization and robustness. We do so by first scaling up the existing detection task setup from the one-generator to multiple-generators in training, during which we disclose two challenges presented in prior methodological designs and demonstrate the divergence of detectors' performance. Specifically, we reveal that the current methods tailored for training on one specific generator either struggle to learn comprehensive artifacts from multiple generators or sacrifice their fitting ability for seen generators (i.e., In-Domain (ID) performance) to exchange the generalization for unseen generators (i.e., Out-Of-Domain (OOD) performance). To tackle the above challenges, we propose our Discrepancy Deepfake Detector (D$^3$) framework, whose core idea is to deconstruct the universal artifacts from multiple generators by introducing a parallel network branch that takes a distorted image feature as an extra discrepancy signal and supplement its original counterpart. Extensive scaled-up experiments demonstrate the effectiveness of D$^3$, achieving 5.3% accuracy improvement in the OOD testing compared to the current SOTA methods while maintaining the ID performance. The source code will be updated in our GitHub repository: https://github.com/BigAandSmallq/D3

cs.CV↗

arXiv2025

Revisiting Classification Taxonomy for Grammatical Errors

Grammatical error classification plays a crucial role in language learning systems, but existing classification taxonomies often lack rigorous validation, leading to inconsistencies and unreliable feedback. In this paper, we revisit previous classification taxonomies for grammatical errors by introducing a systematic and qualitative evaluation framework. Our approach examines four aspects of a taxonomy, i.e., exclusivity, coverage, balance, and usability. Then, we construct a high-quality grammatical error classification dataset annotated with multiple classification taxonomies and evaluate them grounding on our proposed evaluation framework. Our experiments reveal the drawbacks of existing taxonomies. Our contributions aim to improve the precision and effectiveness of error analysis, providing more understandable and actionable feedback for language learners.

cs.CL↗

arXiv2025

Redistribute Ensemble Training for Mitigating Memorization in Diffusion Models

Diffusion models, known for their tremendous ability to generate high-quality samples, have recently raised concerns due to their data memorization behavior, which poses privacy risks. Recent methods for memory mitigation have primarily addressed the issue within the context of the text modality in cross-modal generation tasks, restricting their applicability to specific conditions. In this paper, we propose a novel method for diffusion models from the perspective of visual modality, which is more generic and fundamental for mitigating memorization. Directly exposing visual data to the model increases memorization risk, so we design a framework where models learn through proxy model parameters instead. Specially, the training dataset is divided into multiple shards, with each shard training a proxy model, then aggregated to form the final model. Additionally, practical analysis of training losses illustrates that the losses for easily memorable images tend to be obviously lower. Thus, we skip the samples with abnormally low loss values from the current mini-batch to avoid memorizing. However, balancing the need to skip memorization-prone samples while maintaining sufficient training data for high-quality image generation presents a key challenge. Thus, we propose IET-AGC+, which redistributes highly memorizable samples between shards, to mitigate these samples from over-skipping. Furthermore, we dynamically augment samples based on their loss values to further reduce memorization. Extensive experiments and analysis on four datasets show that our method successfully reduces memory capacity while maintaining performance. Moreover, we fine-tune the pre-trained diffusion models, e.g., Stable Diffusion, and decrease the memorization score by 46.7\%, demonstrating the effectiveness of our method. Code is available in: https://github.com/liuxiao-guan/IET_AGC.

cs.CV↗

arXiv2025

Iterative Ensemble Training with Anti-Gradient Control for Mitigating Memorization in Diffusion Models

Diffusion models, known for their tremendous ability to generate novel and high-quality samples, have recently raised concerns due to their data memorization behavior, which poses privacy risks. Recent approaches for memory mitigation either only focused on the text modality problem in cross-modal generation tasks or utilized data augmentation strategies. In this paper, we propose a novel training framework for diffusion models from the perspective of visual modality, which is more generic and fundamental for mitigating memorization. To facilitate forgetting of stored information in diffusion model parameters, we propose an iterative ensemble training strategy by splitting the data into multiple shards for training multiple models and intermittently aggregating these model parameters. Moreover, practical analysis of losses illustrates that the training loss for easily memorable images tends to be obviously lower. Thus, we propose an anti-gradient control method to exclude the sample with a lower loss value from the current mini-batch to avoid memorizing. Extensive experiments and analysis on four datasets are conducted to illustrate the effectiveness of our method, and results show that our method successfully reduces memory capacity while even improving the performance slightly. Moreover, to save the computing cost, we successfully apply our method to fine-tune the well-trained diffusion models by limited epochs, demonstrating the applicability of our method. Code is available in https://github.com/liuxiao-guan/IET_AGC.

cs.CV↗

arXiv2025

TV-Dialogue: Crafting Theme-Aware Video Dialogues with Immersive Interaction

Recent advancements in LLMs have accelerated the development of dialogue generation across text and images, yet video-based dialogue generation remains underexplored and presents unique challenges. In this paper, we introduce Theme-aware Video Dialogue Crafting (TVDC), a novel task aimed at generating new dialogues that align with video content and adhere to user-specified themes. We propose TV-Dialogue, a novel multi-modal agent framework that ensures both theme alignment (i.e., the dialogue revolves around the theme) and visual consistency (i.e., the dialogue matches the emotions and behaviors of characters in the video) by enabling real-time immersive interactions among video characters, thereby accurately understanding the video content and generating new dialogue that aligns with the given themes. To assess the generated dialogues, we present a multi-granularity evaluation benchmark with high accuracy, interpretability and reliability, demonstrating the effectiveness of TV-Dialogue on self-collected dataset over directly using existing LLMs. Extensive experiments reveal that TV-Dialogue can generate dialogues for videos of any length and any theme in a zero-shot manner without training. Our findings underscore the potential of TV-Dialogue for various applications, such as video re-creation, film dubbing and its use in downstream multimodal tasks.

cs.CV↗

arXiv2025

OCMDP: Observation-Constrained Markov Decision Process

In many practical applications, decision-making processes must balance the costs of acquiring information with the benefits it provides. Traditional control systems often assume full observability, an unrealistic assumption when observations are expensive. We tackle the challenge of simultaneously learning observation and control strategies in such cost-sensitive environments by introducing the Observation-Constrained Markov Decision Process (OCMDP), where the policy influences the observability of the true state. To manage the complexity arising from the combined observation and control actions, we develop an iterative, model-free deep reinforcement learning algorithm that separates the sensing and control components of the policy. This decomposition enables efficient learning in the expanded action space by focusing on when and what to observe, as well as determining optimal control actions, without requiring knowledge of the environment's dynamics. We validate our approach on a simulated diagnostic task and a realistic healthcare environment using HeartPole. Given both scenarios, the experimental results demonstrate that our model achieves a substantial reduction in observation costs on average, significantly outperforming baseline methods by a notable margin in efficiency.

cs.AI↗

arXiv2025

6G Communication New Paradigm: The Integration of Unmanned Aerial Vehicles and Intelligent Reflecting Surfaces

With the continuous development of Intelligent Reflecting Surfaces (IRSs) and Unmanned Aerial Vehicles (UAVs), their combination has become foundational technologies to complement the terrestrial network by providing communication enhancement services for large-scale users. This article provides a comprehensive overview of IRS-assisted UAV communications for 6th-Generation (6G) networks. First, the applications supported by IRS-assisted UAV communications for 6G networks are introduced, and key issues originated from applications supported by IRSs and UAVs for 6G networks are summarized and analyzed. Then, prototypes and main technologies related to the integration of IRSs and UAVs are introduced. Driven by applications and technologies of IRS-assisted UAV communications, existing solutions in the realms of energy-constrained communications, secure communications, and enhanced communications are summarized, and corresponding empirical lessons are provided. Finally, we discuss some research challenges and open issues in IRS-assisted UAV communications, offering directions for the future development.

cs.NI↗

arXiv2025

A self-learning magnetic Hopfield neural network with intrinsic gradient descent adaption

Physical neural networks using physical materials and devices to mimic synapses and neurons offer an energy-efficient way to implement artificial neural networks. Yet, training physical neural networks are difficult and heavily relies on external computing resources. An emerging concept to solve this issue is called physical self-learning that uses intrinsic physical parameters as trainable weights. Under external inputs (i.e. training data), training is achieved by the natural evolution of physical parameters that intrinsically adapt modern learning rules via autonomous physical process, eliminating the requirements on external computation resources.Here, we demonstrate a real spintronic system that mimics Hopfield neural networks (HNN) and unsupervised learning is intrinsically performed via the evolution of physical process. Using magnetic texture defined conductance matrix as trainable weights, we illustrate that under external voltage inputs, the conductance matrix naturally evolves and adapts Oja's learning algorithm in a gradient descent manner. The self-learning HNN is scalable and can achieve associative memories on patterns with high similarities. The fast spin dynamics and reconfigurability of magnetic textures offer an advantageous platform towards efficient autonomous training directly in materials.

cond-mat.dis-nn↗

arXiv2024

Renaissance of Literate Programming in the Era of LLMs: Enhancing LLM-Based Code Generation in Large-Scale Projects

Large Language Models (LLMs) have helped programmers increase efficiency through code generation, comprehension, and repair. However, their application to large-scale projects remains challenging due to complex interdependencies and the extensive size of modern codebases. Although Knuth's concept of Literate Programming (LP) combines code and natural language to convey logic and intent, its potential for enhancing relationships in large projects has not been fully explored. In this study, we introduce the idea of Interoperable LP (ILP), which leverages literate programming principles to enhance the development of both small-scale documents and large-scale projects with LLMs. We investigate how LLMs perform under ILP-style instructions for both document-oriented tasks and entire projects. Recognizing that many researchers rely on well-structured templates to guide LLMs, we propose a concise prompt engineering method to write LP documents so LLMs can better be involved in code generation. We also examine the capacity of various LLMs to generate Scheme and Python code on the RepoBench benchmark, illustrating the advantages of our approach. Our findings indicate that ILP with LLMs can enhance LLM-based code generation in large-scale project development.

cs.SE↗

arXiv2024

On character values of $GL_n(\mathbb F_q)$

In this paper, we use vertex operator techniques to compute character values on unipotent classes of $\GL_n(\mathbb F_q)$. By realizing the Grothendieck ring $R_G=\bigoplus_{n\geq0}^\infty R(\GL_n(\mathbb F_q))$ as Fock spaces, we formulate the Murnanghan-Nakayama rule of $\GL_n(\mathbb F_q)$ between Schur functions colored by an orbit $ϕ$ of linear characters of $\overline{\mathbb F}_q$ under the Frobenius automorphism on and modified Hall-Littlewood functions colored by $f_1=t-1$, which provides detailed information on the character table of $\GL_n(\mathbb F_q)$. As applications, we use vertex algebraic methods to determine the Steinberg characters of $\GL_n(\mathbb F_q)$, which were previously determined by Curtis-Lehrer-Tits via geometry of homology groups of spherical buildings and Springer-Zelevinsky utilizing Hopf algebras.

math.RT↗

arXiv2024

DeepSeek-VL2: Mixture-of-Experts Vision-Language Models for Advanced Multimodal Understanding

We present DeepSeek-VL2, an advanced series of large Mixture-of-Experts (MoE) Vision-Language Models that significantly improves upon its predecessor, DeepSeek-VL, through two key major upgrades. For the vision component, we incorporate a dynamic tiling vision encoding strategy designed for processing high-resolution images with different aspect ratios. For the language component, we leverage DeepSeekMoE models with the Multi-head Latent Attention mechanism, which compresses Key-Value cache into latent vectors, to enable efficient inference and high throughput. Trained on an improved vision-language dataset, DeepSeek-VL2 demonstrates superior capabilities across various tasks, including but not limited to visual question answering, optical character recognition, document/table/chart understanding, and visual grounding. Our model series is composed of three variants: DeepSeek-VL2-Tiny, DeepSeek-VL2-Small and DeepSeek-VL2, with 1.0B, 2.8B and 4.5B activated parameters respectively. DeepSeek-VL2 achieves competitive or state-of-the-art performance with similar or fewer activated parameters compared to existing open-source dense and MoE-based models. Codes and pre-trained models are publicly accessible at https://github.com/deepseek-ai/DeepSeek-VL2.

cs.CV↗

arXiv2024

Rethinking the Intermediate Features in Adversarial Attacks: Misleading Robotic Models via Adversarial Distillation

Language-conditioned robotic learning has significantly enhanced robot adaptability by enabling a single model to execute diverse tasks in response to verbal commands. Despite these advancements, security vulnerabilities within this domain remain largely unexplored. This paper addresses this gap by proposing a novel adversarial prompt attack tailored to language-conditioned robotic models. Our approach involves crafting a universal adversarial prefix that induces the model to perform unintended actions when added to any original prompt. We demonstrate that existing adversarial techniques exhibit limited effectiveness when directly transferred to the robotic domain due to the inherent robustness of discretized robotic action spaces. To overcome this challenge, we propose to optimize adversarial prefixes based on continuous action representations, circumventing the discretization process. Additionally, we identify the beneficial impact of intermediate features on adversarial attacks and leverage the negative gradient of intermediate self-attention features to further enhance attack efficacy. Extensive experiments on VIMA models across 13 robot manipulation tasks validate the superiority of our method over existing approaches and demonstrate its transferability across different model variants.

cs.LG↗

arXiv2024

Weak Host Interactions Induced Thermal Transport Properties of Metal Halide Perovskites Deviating from the Rattling Model

The low-frequency phonon branches of metal halide perovskites typically exhibit the characteristic of hardening with the increase of the cation mass, which leads to anomalous thermal transport phenomenon. However, the underlying physical mechanism is not yet understood. Here, we theoretically compare the thermal transport properties of $A_2$SnI$_6$ ($A$=K, Rb, and Cs) perovskites. The thermal transport in perovskites is widely explained using the rattling model, where ``guest'' cations inside the metal halide framework act as ``rattlers'', but this fails to explain the following phenomenon: The low-frequency phonon branch of $A_2$SnI$_6$ perovskites is insensitive to the mass of the $A^+$ cation and strongly correlated with the interaction of the $A^+$ cation with the I$^-$ anion in the octahedral structures. The failure of the rattling model stems mainly from the weak interactions between the octahedral structures. By developing a new spring model, we successfully explain the thermal transport behavior in $A_2$SnI$_6$ perovskites. Our work gives new insights into the thermal transport mechanism in metal halide perovskites, which has a guiding significance for designing extremely low thermal conductivity materials.

cond-mat.mtrl-sci↗

arXiv2024

Collective Pinning and Vortex Dynamics in type 2 superconducting thin films with Varying Magnetic Field

A perpendicular magnetic field penetrating a thin type-II superconductor slab produces vortices, with one vortex per flux quantum, h/2e. The vortices interact repulsively and form an ordered array (Abrikosov lattice) in clean systems, while strong disorder changes the lattice into a vortex glass. Here we investigate type-II superconducting films (PdBi2 and NbSe2) with surface acoustic waves (SAWs) at mK temperature. When sweeping the magnetic field at an extremely slow rate, we observe a series of spikes in the attenuation and velocity of the SAW, on average separated in field by approximately Hc1. We suspect the following scenario: The vortex-free region at the edges of the film produces an edge barrier across which the vortices can enter or leave. When the applied field changes, the induced supercurrents flowing along this edge region lowers this barrier until there is an instability. At that point, vortices avalanche into (or out of) the bulk and change the vortex crystal, suggested by the sharp jump in each such spike. The vortices then gradually relax to a new stable pinned configuration, leading to a ~30s relaxation after the jump. Our observation enriches the limited experimental evidence on the important topic of real-time vortex dynamics in superconductors.

cond-mat.supr-con↗