arXiv subjects

Yue Xu

Publications and source records attributed to Yue Xu.

At least 55 records · Page 3Linked to original sources

arXiv2025

Deteriorated Interlayer Coupling in Twisted Bilayer Cobaltites

A wealth of remarkable behaviors is observed at the interfaces between magnetic oxides due to the coexistence of Coulomb repulsion and interatomic exchange interactions. While previous research has focused on bonded oxide heterointerfaces, studies on magnetism in van der Waals interfaces remain rare. In this study, we stacked two freestanding cobaltites with precisely controlled twist angles. Scanning transmission electron microscopy revealed clear and ordered moiré patterns, which exhibit an inverse relationship with the twist angle. We found that the Curie temperature in the twisted region is reduced by approximately 13 K compared to the single-layer region using nitrogen-vacancy (NV) magnetometry. This phenomenon may be related to the weakening of the orbital hybridization between oxygen ions and transition metal ions in the unbonded interfaces. Our findings suggest a potential avenue for modulating magnetic interactions in correlated systems through twist, providing opportunities for the discovery of unknown quantum states.

cond-mat.mtrl-sci↗

arXiv2024

Dense halves in balanced 2-partition of K4-free graphs

A balanced 2-partition of a graph is a bipartition $A,A^c$ of $V(G)$ such that $|A|=|A^c|$. Balogh, Clemen, and Lidický conjectured that for every $K_4$-free graph on $n$ (even) vertices, there exists a balanced 2-partition $A,A^c$ such that $\max\{e(A),e(A^c)\}\leq n^2/16$ edges. In this paper, we present a family of counterexamples to the conjecture and provide a new upper bound ($0.074n^2$) for every sufficiently large even integer $n$.

math.CO↗

arXiv2024

CARES: A Comprehensive Benchmark of Trustworthiness in Medical Vision Language Models

Artificial intelligence has significantly impacted medical applications, particularly with the advent of Medical Large Vision Language Models (Med-LVLMs), sparking optimism for the future of automated and personalized healthcare. However, the trustworthiness of Med-LVLMs remains unverified, posing significant risks for future model deployment. In this paper, we introduce CARES and aim to comprehensively evaluate the Trustworthiness of Med-LVLMs across the medical domain. We assess the trustworthiness of Med-LVLMs across five dimensions, including trustfulness, fairness, safety, privacy, and robustness. CARES comprises about 41K question-answer pairs in both closed and open-ended formats, covering 16 medical image modalities and 27 anatomical regions. Our analysis reveals that the models consistently exhibit concerns regarding trustworthiness, often displaying factual inaccuracies and failing to maintain fairness across different demographic groups. Furthermore, they are vulnerable to attacks and demonstrate a lack of privacy awareness. We publicly release our benchmark and code in https://cares-ai.github.io/.

cs.LG↗

arXiv2024

$\textit{MMJ-Bench}$: A Comprehensive Study on Jailbreak Attacks and Defenses for Multimodal Large Language Models

As deep learning advances, Large Language Models (LLMs) and their multimodal counterparts, Multimodal Large Language Models (MLLMs), have shown exceptional performance in many real-world tasks. However, MLLMs face significant security challenges, such as jailbreak attacks, where attackers attempt to bypass the model's safety alignment to elicit harmful responses. The threat of jailbreak attacks on MLLMs arises from both the inherent vulnerabilities of LLMs and the multiple information channels that MLLMs process. While various attacks and defenses have been proposed, there is a notable gap in unified and comprehensive evaluations, as each method is evaluated on different dataset and metrics, making it impossible to compare the effectiveness of each method. To address this gap, we introduce \textit{MMJ-Bench}, a unified pipeline for evaluating jailbreak attacks and defense techniques for MLLMs. Through extensive experiments, we assess the effectiveness of various attack methods against SoTA MLLMs and evaluate the impact of defense mechanisms on both defense effectiveness and model utility for normal tasks. Our comprehensive evaluation contribute to the field by offering a unified and systematic evaluation framework and the first public-available benchmark for MLLM jailbreak research. We also demonstrate several insightful findings that highlights directions for future studies.

cs.CR↗

arXiv2024

Graph Neural Patching for Cold-Start Recommendations

The cold start problem in recommender systems remains a critical challenge. Current solutions often train hybrid models on auxiliary data for both cold and warm users/items, potentially degrading the experience for the latter. This drawback limits their viability in practical scenarios where the satisfaction of existing warm users/items is paramount. Although graph neural networks (GNNs) excel at warm recommendations by effective collaborative signal modeling, they haven't been effectively leveraged for the cold-start issue within a user-item graph, which is largely due to the lack of initial connections for cold user/item entities. Addressing this requires a GNN adept at cold-start recommendations without sacrificing performance for existing ones. To this end, we introduce Graph Neural Patching for Cold-Start Recommendations (GNP), a customized GNN framework with dual functionalities: GWarmer for modeling collaborative signal on existing warm users/items and Patching Networks for simulating and enhancing GWarmer's performance on cold-start recommendations. Extensive experiments on three benchmark datasets confirm GNP's superiority in recommending both warm and cold users/items.

cs.IR↗

arXiv2024

Cross-modality Information Check for Detecting Jailbreaking in Multimodal Large Language Models

Multimodal Large Language Models (MLLMs) extend the capacity of LLMs to understand multimodal information comprehensively, achieving remarkable performance in many vision-centric tasks. Despite that, recent studies have shown that these models are susceptible to jailbreak attacks, which refer to an exploitative technique where malicious users can break the safety alignment of the target model and generate misleading and harmful answers. This potential threat is caused by both the inherent vulnerabilities of LLM and the larger attack scope introduced by vision input. To enhance the security of MLLMs against jailbreak attacks, researchers have developed various defense techniques. However, these methods either require modifications to the model's internal structure or demand significant computational resources during the inference phase. Multimodal information is a double-edged sword. While it increases the risk of attacks, it also provides additional data that can enhance safeguards. Inspired by this, we propose Cross-modality Information DEtectoR (CIDER), a plug-and-play jailbreaking detector designed to identify maliciously perturbed image inputs, utilizing the cross-modal similarity between harmful queries and adversarial images. CIDER is independent of the target MLLMs and requires less computation cost. Extensive experimental results demonstrate the effectiveness and efficiency of CIDER, as well as its transferability to both white-box and black-box MLLMs.

cs.CL↗

arXiv2024

Developing Guidelines for Functionally-Grounded Evaluation of Explainable Artificial Intelligence using Tabular Data

Explainable Artificial Intelligence (XAI) techniques are used to provide transparency to complex, opaque predictive models. However, these techniques are often designed for image and text data, and it is unclear how fit-for-purpose they are when applied to tabular data. As XAI techniques are rarely evaluated in settings with tabular data, the applicability of existing evaluation criteria and methods are also unclear and needs (re-)examination. For example, some works suggest that evaluation methods may unduly influence the evaluation results when using tabular data. This lack of clarity on evaluation procedures can lead to reduced transparency and ineffective use of XAI techniques in real world settings. In this study, we examine literature on XAI evaluation to derive guidelines on functionally-grounded assessment of local, post hoc XAI techniques. We identify 20 evaluation criteria and associated evaluation methods, and derive guidelines on when and how each criterion should be evaluated. We also identify key research gaps to be addressed by future work. Our study contributes to the body of knowledge on XAI evaluation through in-depth examination of functionally-grounded XAI evaluation protocols, and has laid the groundwork for future research on XAI evaluation.

cs.CY↗

arXiv2024

Leveraging Large Language Models for Generating Mobile Sensing Strategies in Human Behavior Modeling

Mobile sensing plays a crucial role in generating digital traces to understand human daily lives. However, studying behaviours like mood or sleep quality in smartphone users requires carefully designed mobile sensing strategies such as sensor selection and feature construction. This process is time-consuming, burdensome, and requires expertise in multiple domains. Furthermore, the resulting sensing framework lacks generalizability, making it difficult to apply to different scenarios. In the research, we propose an automated mobile sensing strategy for human behaviour understanding. First, we establish a knowledge base and consolidate rules for data collection and effective feature construction. Then, we introduce the multi-granular human behaviour representation and design procedures for leveraging large language models to generate strategies. Our approach is validated through blind comparative studies and usability evaluation. Ultimately, our approach holds the potential to revolutionise the field of mobile sensing and its applications.

cs.HC↗

arXiv2024

Distill Gold from Massive Ores: Bi-level Data Pruning towards Efficient Dataset Distillation

Data-efficient learning has garnered significant attention, especially given the current trend of large multi-modal models. Recently, dataset distillation has become an effective approach by synthesizing data samples that are essential for network training. However, it remains to be explored which samples are essential for the dataset distillation process itself. In this work, we study the data efficiency and selection for the dataset distillation task. By re-formulating the dynamics of distillation, we provide insight into the inherent redundancy in the real dataset, both theoretically and empirically. We propose to use the empirical loss value as a static data pruning criterion. To further compensate for the variation of the data value in training, we find the most contributing samples based on their causal effects on the distillation. The proposed selection strategy can efficiently exploit the training dataset, outperform the previous SOTA distillation algorithms, and consistently enhance the distillation algorithms, even on much larger-scale and more heterogeneous datasets, e.g., full ImageNet-1K and Kinetics-400. We believe this paradigm will open up new avenues in the dynamics of distillation and pave the way for efficient dataset distillation. Our code is available on https://github.com/silicx/GoldFromOres-BiLP.

cs.LG↗

arXiv2024

Take A Step Back: Rethinking the Two Stages in Visual Reasoning

Visual reasoning, as a prominent research area, plays a crucial role in AI by facilitating concept formation and interaction with the world. However, current works are usually carried out separately on small datasets thus lacking generalization ability. Through rigorous evaluation of diverse benchmarks, we demonstrate the shortcomings of existing ad-hoc methods in achieving cross-domain reasoning and their tendency to data bias fitting. In this paper, we revisit visual reasoning with a two-stage perspective: (1) symbolization and (2) logical reasoning given symbols or their representations. We find that the reasoning stage is better at generalization than symbolization. Thus, it is more efficient to implement symbolization via separated encoders for different data domains while using a shared reasoner. Given our findings, we establish design principles for visual reasoning frameworks following the separated symbolization and shared reasoning. The proposed two-stage framework achieves impressive generalization ability on various visual reasoning tasks, including puzzles, physical prediction, and visual question answering (VQA), encompassing both 2D and 3D modalities. We believe our insights will pave the way for generalizable visual reasoning.

cs.CV↗

arXiv2024

FAFE: Immune Complex Modeling with Geodesic Distance Loss on Noisy Group Frames

Despite the striking success of general protein folding models such as AlphaFold2(AF2, Jumper et al. (2021)), the accurate computational modeling of antibody-antigen complexes remains a challenging task. In this paper, we first analyze AF2's primary loss function, known as the Frame Aligned Point Error (FAPE), and raise a previously overlooked issue that FAPE tends to face gradient vanishing problem on high-rotational-error targets. To address this fundamental limitation, we propose a novel geodesic loss called Frame Aligned Frame Error (FAFE, denoted as F2E to distinguish from FAPE), which enables the model to better optimize both the rotational and translational errors between two frames. We then prove that F2E can be reformulated as a group-aware geodesic loss, which translates the optimization of the residue-to-residue error to optimizing group-to-group geodesic frame distance. By fine-tuning AF2 with our proposed new loss function, we attain a correct rate of 52.3\% (DockQ $>$ 0.23) on an evaluation set and 43.8\% correct rate on a subset with low homology, with substantial improvement over AF2 by 182\% and 100\% respectively.

q-bio.QM↗

arXiv2024

Dual Attribute-Spatial Relation Alignment for 3D Visual Grounding

3D visual grounding is an emerging research area dedicated to making connections between the 3D physical world and natural language, which is crucial for achieving embodied intelligence. In this paper, we propose DASANet, a Dual Attribute-Spatial relation Alignment Network that separately models and aligns object attributes and spatial relation features between language and 3D vision modalities. We decompose both the language and 3D point cloud input into two separate parts and design a dual-branch attention module to separately model the decomposed inputs while preserving global context in attribute-spatial feature fusion by cross attentions. Our DASANet achieves the highest grounding accuracy 65.1% on the Nr3D dataset, 1.3% higher than the best competitor. Besides, the visualization of the two branches proves that our method is efficient and highly interpretable.

cs.CV↗

arXiv2024

Low-Rank Similarity Mining for Multimodal Dataset Distillation

Though dataset distillation has witnessed rapid development in recent years, the distillation of multimodal data, e.g., image-text pairs, poses unique and under-explored challenges. Unlike unimodal data, image-text contrastive learning (ITC) data lack inherent categorization and should instead place greater emphasis on modality correspondence. In this work, we propose Low-Rank Similarity Mining (LoRS) for multimodal dataset distillation, that concurrently distills a ground truth similarity matrix with image-text pairs, and leverages low-rank factorization for efficiency and scalability. The proposed approach brings significant improvement to the existing algorithms, marking a significant contribution to the field of visual-language dataset distillation. We advocate adopting LoRS as a foundational synthetic data setup for image-text dataset distillation. Our code is available at https://github.com/silicx/LoRS_Distill.

cs.LG↗

arXiv2024

Protective Effect of Trehalose Sugar on Amyloid-Membrane Interactions using BLM Electrophysiology

Alzheimer's disease (AD) is a neurodegenerative disease characterized by dementia and memory loss in the elderly population. The amyloid-$β$ peptide (A$β$ 1-42) is one of the main pathogenic factors in AD and is known to cause damage to neuronal cellular membranes. There is no cure currently available for AD and new approaches, including preventive strategies, are highly desirable. In this work, we explore the possibility of protecting neuronal membranes from amyloid-induced damage with naturally existing sugar trehalose. Trehalose has been shown to protect plant cellular membranes in extreme conditions and modify A$β$ misfolding. We hypothesize that trehalose can protect the neuronal membrane from amyloid toxicity. In this work, we studied the protective effect of trehalose against A$β$-induced damage in model lipid membranes (DPPC/POPC/Cholesterol) to mimic neuronal membranes using atomic force microscopy (AFM) and black lipid membrane (BLM) electrophysiology. Our results demonstrated that A$β$1-42 damaged membranes and led to ionic current leakage across these membranes due to the formation of various defects and pores. The presence of trehalose reduced the ion current across membranes caused by A$β$1-42 peptide damage, thus efficiently protecting the membranes. These findings suggest the trehalose sugar can be useful in protecting neuronal cellular membranes against amyloid toxicity in AD prevention.

physics.bio-ph↗

arXiv2024

Macro Graph Neural Networks for Online Billion-Scale Recommender Systems

Predicting Click-Through Rate (CTR) in billion-scale recommender systems poses a long-standing challenge for Graph Neural Networks (GNNs) due to the overwhelming computational complexity involved in aggregating billions of neighbors. To tackle this, GNN-based CTR models usually sample hundreds of neighbors out of the billions to facilitate efficient online recommendations. However, sampling only a small portion of neighbors results in a severe sampling bias and the failure to encompass the full spectrum of user or item behavioral patterns. To address this challenge, we name the conventional user-item recommendation graph as "micro recommendation graph" and introduce a more suitable MAcro Recommendation Graph (MAG) for billion-scale recommendations. MAG resolves the computational complexity problems in the infrastructure by reducing the node count from billions to hundreds. Specifically, MAG groups micro nodes (users and items) with similar behavior patterns to form macro nodes. Subsequently, we introduce tailored Macro Graph Neural Networks (MacGNN) to aggregate information on a macro level and revise the embeddings of macro nodes. MacGNN has already served Taobao's homepage feed for two months, providing recommendations for over one billion users. Extensive offline experiments on three public benchmark datasets and an industrial dataset present that MacGNN significantly outperforms twelve CTR baselines while remaining computationally efficient. Besides, online A/B tests confirm MacGNN's superiority in billion-scale recommender systems.

cs.IR↗

arXiv2024

Optically Detected Magnetic Resonance of Nitrogen-Vacancy Centers in Diamond under Weak Laser Excitation

As promising quantum sensors, nitrogen-vacancy (NV) centers in diamond have been widely used in frontier studies in condensed matter physics, material sciences, and life sciences. In practical applications, weak laser excitation is favorable as it reduces the side effects of laser irradiation, for example, phototoxicity and heating. Here we report a combined theoretical and experimental study of optically detected magnetic resonance (ODMR) of NV-center ensembles under weak 532-nm laser excitation. In this regime, both the width and splitting of ODMR spectra decrease with increasing laser power. This power dependence is reproduced with a model considering laser-induced charge neutralization of NV--N+ pairs, which alters the local electric field environment. These results are important for understanding and designing NV-based quantum sensing in light-sensitive applications.

quant-ph↗

arXiv2024

Dancing with Still Images: Video Distillation via Static-Dynamic Disentanglement

Recently, dataset distillation has paved the way towards efficient machine learning, especially for image datasets. However, the distillation for videos, characterized by an exclusive temporal dimension, remains an underexplored domain. In this work, we provide the first systematic study of video distillation and introduce a taxonomy to categorize temporal compression. Our investigation reveals that the temporal information is usually not well learned during distillation, and the temporal dimension of synthetic data contributes little. The observations motivate our unified framework of disentangling the dynamic and static information in the videos. It first distills the videos into still images as static memory and then compensates the dynamic and motion information with a learnable dynamic memory block. Our method achieves state-of-the-art on video datasets at different scales, with a notably smaller memory storage budget. Our code is available at https://github.com/yuz1wan/video_distillation.

cs.CV↗

arXiv2024

Quantum Assisted Stochastic Economic Dispatch for Renewables Rich Power Systems

Considering widely dispersed uncertain renewable energy sources (RESs), scenario-based stochastic optimization is an effective method for the economic dispatch of renewables-rich power systems. However, on classic computers, to simulate RES uncertainties with high accuracy, the massive scenario generation is very time-consuming, and the pertinent optimization problem is high-dimensional NP-hard mixed-integer programming. To this end, we design a quantum-assisted scheme to accelerate the stochastic optimization for power system economic dispatch without losing accuracy. We first propose the unified quantum amplitude estimation to characterize RES uncertainties, thereby generating massive scenarios by a few qubits to reduce state variables. Then, strong Benders cuts corresponding to some specific scenarios are selected to control the solution scale of Benders master problem in the iterative process, all of which are implemented by customized quantum approximation optimization algorithms. Finally, we perform numerical experiments on the modified IEEE 6-bus system to test the designed scheme.

math.OC↗