arXiv subjects

Xiaomeng Wang

Publications and source records attributed to Xiaomeng Wang.

At least 19 recordsLinked to original sources

arXiv2026

Visual Input and Its Framing Affect Attribute-based Descriptions Produced by Large Vision-Language Models

Large vision-language models (LVLMs) are commonly used with only a single text prompt as the input, or plus an image. In this paper, we demonstrate that when the image exists, even if the text prompt is not about the specific instance (but only the concept it belongs to) in that image, the response would still be affected. For example, when the text prompt only asks for the attribute descriptions of a dog breed, an image depicting a specific dog from that breed would shift the response. Further, how the specific instance is framed in that image would determine towards which the response shifts. Detailed analyses also reveal that in the response, physical terms increase from 18% for text-only to 45% (40%) for subject-focused (subject-in-situation) framings. Overall, the unexpected effects of visual cues on LVLMs highlight the need to understand the presence of an image and its framing when evaluating the robustness of LVLMs.

cs.CV↗

arXiv2026

Triggering Stealthy Feature Map Backdoors via Physical Fault Injection in Embedded Neural Networks

Fault injection (FI) attacks on embedded neural network (NN) implementations primarily focus on inducing misclassification by corrupting weights or intermediate computations, overlooking their interaction with algorithmic adversarial threats. In this work, we present a cross-level attack that bridges implementation-level physical faults to algorithm-level adversarial attacks. By characterizing fault-induced data perturbations during NN inference, we connect FI with backdoor learning, enabling system-level attacks that jointly exploit implementation- and algorithm-level vulnerabilities. Specifically, we propose a precise fault-injection method that reliably manipulates targeted register values to tractable states during execution. Leveraging this level of FI precision, we propose a novel end-to-end feature map-level backdoor attack, where physically induced intermediate perturbations serve as stealthy triggers. Unlike conventional input-based backdoors, our trigger is activated only under physical faults, causing the NN to exhibit adversarial behavior that compromises system integrity while remaining benign during normal operation. We demonstrate that such physically triggered backdoors can be mounted on embedded NN platforms and remain effective against existing backdoor defenses that typically assume input-space triggers. We showcase the attack practicality using electromagnetic FI on convolutional neural networks implemented on ARM Cortex-M4 microcontroller, which is a common platform for constrained embedded applications. Our results highlight a novel attack vector at the intersection of hardware and algorithmic levels, stressing the need for defenses across abstraction levels.

cs.CR↗

arXiv2026

TRACE: A Temporal Conditional Estimation for Multimodal Time Series Foundation Models

Time series foundation models (TS-FMs) aim to learn generalizable temporal representations that can be adapted to a wide range of downstream tasks. In real-world multimodal settings, time series are frequently affected by temporal misalignment and partial modality missingness, where different modalities are observed at heterogeneous time scales or are partially absent. Existing approaches typically rely on naive imputation or masking strategies, which fail to account for cross-modal dependencies and often lead to misaligned or degraded representations. We propose TRACE, a conditional estimation paradigm for multimodal time series foundation model pipelines under missingness and irregular sampling, allowing incomplete target modalities to be systematically inferred from available auxiliary modalities. We evaluate TRACE on diverse multimodal benchmarks spanning healthcare and affective computing, including the MIMIC-IV clinical dataset and the CMU-MOSI and CMU-MOSEI benchmarks for multimodal sentiment analysis. Across a range of downstream prediction tasks and missing-modality settings, TRACE consistently outperforms prior multimodal fusion approaches, demonstrating improved robustness to severe modality missingness and more reliable cross-modal representations.

cs.AI↗

arXiv2026

Stability of nontrivial graph pairs

A graph pair $(Γ, Σ)$ is called stable if every automorphism of the direct product $Γ\timesΣ$ is induced componentwise by automorphisms of $Γ$ and $Σ$. A graph is twin-free if no two distinct vertices share the same neighbourhood in the graph. Two graphs $Γ$ and $Σ$ are coprime with respect to the direct product if there is no graph $Δ$ of order greater than $1$ such that $Γ\congΓ'\timesΔ$ and $Σ\congΣ'\timesΔ$ for some graphs $Γ'$ and $Σ'$. A graph pair $(Γ,Σ)$ is nontrivial if $Γ$ and $Σ$ are coprime connected twin-free graphs and exactly one of them is bipartite. In this paper, we prove that if $Γ$ is non-bipartite, stable, and factor-loopless, then each nontrivial graph pair $(Γ,Σ)$ is stable. This gives a partial answer to [Question~19, Qin, Xia and Zhou, Discrete Math., 113856, (2024)] and proves the factor-loopless case of [Conjecture~1.3, Wang, Qin and Xia, arXiv:2509.26170]. We also give affirmative answers to [Questions~3.5, 3.6, Gan, Liu and Xia, J. Combin. Theory Ser. B, 140--164, (2025)] and a negative answer to [Question~3.7, Gan, Liu and Xia, J. Combin. Theory Ser. B, 140--164, (2025)].

math.CO↗

arXiv2026

Learning to target with network interference

This paper studies adaptive targeting under network interference in a bandit setting, where treatments applied to one individual may affect others through spillover effects. We consider a linear model in a sparse regime, where each individual's outcome can be affected by at most a few others. We first establish a regret lower bound showing that ignoring the network structure and reducing the problem to a standard linear bandit inevitably leads to inefficient learning, particularly in large populations. To understand how structural information can be leveraged, we analyze regimes with varying levels of knowledge of the interference structure: (1) full support knowledge, (2) knowledge of the column support sizes, and (3) no prior knowledge. For each regime, we establish regret lower bounds characterizing the fundamental limits of learning, and develop algorithms that achieve near-optimal regret. Together, our results provide a unified view of how knowledge of the interference structure governs the efficiency of online learning under interference, and offer practical adaptive targeting algorithms in each setting. Numerical experiments on synthetic and real-world data demonstrate the practical benefits of our algorithms.

stat.ML↗

arXiv2026

The existence of unexpected automorphisms in direct product graphs

A pair of graphs $(Γ,Σ)$ is called unstable if their direct product $Γ\timesΣ$ admits automorphisms not from $\mathrm{Aut}(Γ)\times\mathrm{Aut}(Σ)$, and such automorphisms are said to be unexpected. The stability of a graph $Γ$ refers to that of $(Γ,K_2)$. While the stability of individual graphs has been relatively well studied, much less is known for graph pairs. In this paper, we propose a conjecture that provides the best possible reduction of the stability of a graph pair to the stability of a single graph. We prove one direction of this conjecture and establish partial results for the converse. This enables the determination of the stability of a broad class of graph pairs, with complete results when one factor is a cycle.

math.CO↗

arXiv2026

Revealing the Impact of Visual Text Style on Attribute-based Descriptions Produced by Large Visual Language Models

When the visual style of text is considered, a wide variety can be observed in font, color, and size. However, when a word is read, its meaning is independent of the style in which it has been written or rendered. In this paper, we investigate whether, and how, the style in which a word is visualized in an image impacts the description that a Large Visual Language Model (LVLM) provides for the concept to which that word refers. Specifically, we investigate how functional text styles (readability-oriented, e.g., black sans-serif) versus decorative styles (display-oriented, e.g., colored cursive/script) affect LVLMs' descriptions of a concept in terms of the attributes of that concept. Our experiments study the situation in which the LVLM is able to correctly identify the concept referred to by a visual text, i.e., by a word or words rendered as an image, and in which the visual text style should not influence the attribute-based description that the LVLM produces. Our experimental results reveal that even when the concept is correctly identified, text style influences the model's attribute-based descriptions of the concept. Our findings demonstrate non-trivial style leakage from text style into semantic inference and motivate style-aware evaluation and mitigation for LVLM-based multimedia systems.

cs.CV↗

arXiv2026

Classification of perfect and total perfect codes in generalized Petersen graphs

In a graph $Γ$, a perfect code is an independent set $C$ with the property that every vertex not in $C$ is adjacent to a unique vertex in $C$, and a total perfect code is a set $C$ of vertices of $Γ$ such that every vertex of $Γ$ is adjacent to a unique vertex in $C$. We classify these codes for generalized Petersen graphs.

math.CO↗

arXiv2026

DS-CIM: Digital Stochastic Computing-In-Memory Featuring Accurate OR-Accumulation via Sample Region Remapping for Edge AI Models

Stochastic computing (SC) offers hardware simplicity but suffers from low throughput, while high-throughput Digital Computing-in-Memory (DCIM) is bottlenecked by costly adder logic for matrix-vector multiplication (MVM). To address this trade-off, this paper introduces a digital stochastic CIM (DS-CIM) architecture that achieves both high accuracy and efficiency. We implement signed multiply-accumulation (MAC) in a compact, unsigned OR-based circuit by modifying the data representation. Throughput is enhanced by replicating this low-cost circuit 64 times with only a 1x area increase. Our core strategy, a shared Pseudo Random Number Generator (PRNG) with 2D partitioning, enables single-cycle mutually exclusive activation to eliminate OR-gate collisions. We also resolve the 1s saturation issue via stochastic process analysis and data remapping, significantly improving accuracy and resilience to input sparsity. Our high-accuracy DS-CIM1 variant achieves 94.45% accuracy for INT8 ResNet18 on CIFAR-10 with a root-mean-squared error (RMSE) of just 0.74%. Meanwhile, our high-efficiency DS-CIM2 variant attains an energy efficiency of 3566.1 TOPS/W and an area efficiency of 363.7 TOPS/mm^2, while maintaining a low RMSE of 3.81%. The DS-CIM capability with larger models is further demonstrated through experiments with INT8 ResNet50 on ImageNet and the FP8 LLaMA-7B model.

cs.AR↗

arXiv2025

DIRC-RAG: Accelerating Edge RAG with Robust High-Density and High-Loading-Bandwidth Digital In-ReRAM Computation

Retrieval-Augmented Generation (RAG) enhances large language models (LLMs) by integrating external knowledge retrieval but faces challenges on edge devices due to high storage, energy, and latency demands. Computing-in-Memory (CIM) offers a promising solution by storing document embeddings in CIM macros and enabling in-situ parallel retrievals but is constrained by either low memory density or limited computational accuracy. To address these challenges, we present DIRCRAG, a novel edge RAG acceleration architecture leveraging Digital In-ReRAM Computation (DIRC). DIRC integrates a high-density multi-level ReRAM subarray with an SRAM cell, utilizing SRAM and differential sensing for robust ReRAM readout and digital multiply-accumulate (MAC) operations. By storing all document embeddings within the CIM macro, DIRC achieves ultra-low-power, single-cycle data loading, substantially reducing both energy consumption and latency compared to offchip DRAM. A query-stationary (QS) dataflow is supported for RAG tasks, minimizing on-chip data movement and reducing SRAM buffer requirements. We introduce error optimization for the DIRC ReRAM-SRAM cell by extracting the bit-wise spatial error distribution of the ReRAM subarray and applying targeted bit-wise data remapping. An error detection circuit is also implemented to enhance readout resilience against deviceand circuit-level variations. Simulation results demonstrate that DIRC-RAG under TSMC40nm process achieves an on-chip non-volatile memory density of 5.18Mb/mm2 and a throughput of 131 TOPS. It delivers a 4MB retrieval latency of 5.6μs/query and an energy consumption of 0.956μJ/query, while maintaining the retrieval precision.

cs.AR↗

arXiv2025

Signature of gate tunable superconducting network in twisted bilayer graphene

Twisted van der Waals materials provide a tunable platform for investigating two-dimensional superconductivity and quantum phases. Using spectra-imaging scanning tunneling microscopy, we study the superconducting states in twisted bilayer graphene and track their evolution from insulating phases. Gate-dependent spectroscopic measurements reveal two distinct regimes: under-doped (ν = -2.3) and optimally doped (ν = -2.6). In the under-doped regime, partial superconductivity arises, forming a network interspersed with non-gapped regions. At optimal doping, the entire unit cell demonstrates superconductivity, with gap size modulation showing an anti-correlation with the local density of states. This gate-dependent transition from an insulating phase to a modulated superconductor uncovers an unexpected spatial hierarchy in pairing behavior and offers direct microscopic insights to constrain theories of superconductivity in moiré systems.

cond-mat.supr-con↗

arXiv2025

Fast Rerandomization via the BRAIN

Randomized experiments are a crucial tool for causal inference in many different fields. Rerandomization addresses any covariate imbalance in such experiments by resampling treatment assignments until certain balance criteria are satisfied. However, rerandomization based on naïve acceptance-rejection sampling is computationally inefficient, especially when numerous independent assignments are required to perform randomization-based statistical inference. Existing acceleration methods are suboptimal and not applicable in structured experiments, including stratified and clustered experiments. Based on metaheuristics in integer programming, we propose BRAIN -- a novel computationally-lightweight methodology that ensures covariate balance in randomized experiments while significantly accelerating the computation. Our BRAIN method provides unbiased treatment effect estimators with reduced variance compared to complete randomization, preserving the desirable statistical properties of traditional rerandomization. Simulation studies and a real data example demonstrate the benefits of our method in fast sampling while retaining the appealing statistical guarantees.

stat.ME↗

arXiv2025

GeoGrid-Bench: Can Foundation Models Understand Multimodal Gridded Geo-Spatial Data?

We present GeoGrid-Bench, a benchmark designed to evaluate the ability of foundation models to understand geo-spatial data in the grid structure. Geo-spatial datasets pose distinct challenges due to their dense numerical values, strong spatial and temporal dependencies, and unique multimodal representations including tabular data, heatmaps, and geographic visualizations. To assess how foundation models can support scientific research in this domain, GeoGrid-Bench features large-scale, real-world data covering 16 climate variables across 150 locations and extended time frames. The benchmark includes approximately 3,200 question-answer pairs, systematically generated from 8 domain expert-curated templates to reflect practical tasks encountered by human scientists. These range from basic queries at a single location and time to complex spatiotemporal comparisons across regions and periods. Our evaluation reveals that vision-language models perform best overall, and we provide a fine-grained analysis of the strengths and limitations of different foundation models in different geo-spatial tasks. This benchmark offers clearer insights into how foundation models can be effectively applied to geo-spatial data analysis and used to support scientific research.

cs.CL↗

arXiv2025

Partial Knowledge Distillation for Alleviating the Inherent Inter-Class Discrepancy in Federated Learning

Substantial efforts have been devoted to alleviating the impact of the long-tailed class distribution in federated learning. In this work, we observe an interesting phenomenon that certain weak classes consistently exist even for class-balanced learning. These weak classes, different from the minority classes in the previous works, are inherent to data and remain fairly consistent for various network structures, learning paradigms, and data partitioning methods. The inherent inter-class accuracy discrepancy can reach over 36.9% for federated learning on the FashionMNIST and CIFAR-10 datasets, even when the class distribution is balanced both globally and locally. In this study, we empirically analyze the potential reason for this phenomenon. Furthermore, a partial knowledge distillation (PKD) method is proposed to improve the model's classification accuracy for weak classes. In this approach, knowledge transfer is initiated upon the occurrence of specific misclassifications within certain weak classes. Experimental results show that the accuracy of weak classes can be improved by 10.7%, reducing the inherent inter-class discrepancy effectively.

cs.LG↗

arXiv2025

StarGen: A Spatiotemporal Autoregression Framework with Video Diffusion Model for Scalable and Controllable Scene Generation

Recent advances in large reconstruction and generative models have significantly improved scene reconstruction and novel view generation. However, due to compute limitations, each inference with these large models is confined to a small area, making long-range consistent scene generation challenging. To address this, we propose StarGen, a novel framework that employs a pre-trained video diffusion model in an autoregressive manner for long-range scene generation. The generation of each video clip is conditioned on the 3D warping of spatially adjacent images and the temporally overlapping image from previously generated clips, improving spatiotemporal consistency in long-range scene generation with precise pose control. The spatiotemporal condition is compatible with various input conditions, facilitating diverse tasks, including sparse view interpolation, perpetual view generation, and layout-conditioned city generation. Quantitative and qualitative evaluations demonstrate StarGen's superior scalability, fidelity, and pose accuracy compared to state-of-the-art methods. Project page: https://zju3dv.github.io/StarGen.

cs.CV↗

arXiv2025

Data-driven Camera and Lidar Simulation Models for Autonomous Driving: A Review from Generative Models to Volume Renderers

Perception sensors, particularly camera and Lidar, are key elements of Autonomous Driving Systems (ADS) that enable them to comprehend their surroundings to informed driving and control decisions. Therefore, developing realistic simulation models for these sensors is essential for conducting effective simulation-based testing of ADS. Moreover, the rise of deep learning-based perception models has increased the utility of sensor simulation models for synthesising diverse training datasets. The traditional sensor simulation models rely on computationally expensive physics-based algorithms, specifically in complex systems such as ADS. Hence, the current potential resides in data-driven approaches, fuelled by the exceptional performance of deep generative models in capturing high-dimensional data distribution and volume renderers in accurately representing scenes. This paper reviews the current state-of-the-art data-driven camera and Lidar simulation models and their evaluation methods. It explores a spectrum of models from the novel perspective of generative models and volume renderers. Generative models are discussed in terms of their input-output types, while volume renderers are categorised based on their input encoding. Finally, the paper illustrates commonly used evaluation techniques for assessing sensor simulation models and highlights the existing research gaps in the area.

cs.CV↗

arXiv2025

Spectral signature of periodic modulation and sliding of pseudogap state in moire system

The nature of the pseudogap state is widely believed as a key to understanding the pairing mechanism underlying unconventional superconductivity. Over the past two decades, significant efforts have been devoted to searching for spontaneous symmetry breaking or potential order parameters associated with these pseudogap states, aiming to better characterize their properties. Recently, pseudogap states have also been realized in moire systems with extensive gate tunability, yet their local electronic structure remains largely unexplored8. In this study, we report the observation of gate-tunable spontaneous symmetry breaking and sliding behavior of the pseudogap state in magic-angle twisted bilayer graphene (MAtBG) using spectroscopic imaging scanning tunneling microscopy. Our spectroscopy reveals a distinct pseudogap at 4.4 K within the doping range -3 < v < -2. Spectroscopic imaging highlights a gap size modulation at moire scale that is sensitive to the filling, indicative of a wave-like fluctuating pseudogap feature. Specifically, the positions of gap size minima (GSM) coincide with regions of the highest local density of states (LDOS) at the filling v = -2.63, but a unidirectional sliding behavior of GSM is observed for other fillings. In addition, the pseudogap size distribution at certain doping levels also causes a clear nematic order, or an anisotropic gap distribution. Our results have shed light on the complex nature of this pseudogap state, revealing critical insights into the phase diagram of correlated electron systems.

cond-mat.mes-hall↗

arXiv2025

Towards Rationality in Language and Multimodal Agents: A Survey

This work discusses how to build more rational language and multimodal agents and what criteria define rationality in intelligent systems. Rationality is the quality of being guided by reason, characterized by decision-making that aligns with evidence and logical principles. It plays a crucial role in reliable problem-solving by ensuring well-grounded and consistent solutions. Despite their progress, large language models (LLMs) often fall short of rationality due to their bounded knowledge space and inconsistent outputs. In response, recent efforts have shifted toward developing multimodal and multi-agent systems, as well as integrating modules like external tools, programming codes, symbolic reasoners, utility function, and conformal risk controls rather than relying solely on a single LLM for decision-making. This paper surveys state-of-the-art advancements in language and multimodal agents, assesses their role in enhancing rationality, and outlines open challenges and future research directions. We maintain an open repository at https://github.com/bowen-upenn/Agent_Rationality.

cs.AI↗