arXiv subjects

Xin Zhou

Publications and source records attributed to Xin Zhou.

At least 127 records · Page 7Linked to original sources

arXiv2025

Molecular Gas Distribution toward the Inner and Outer Galaxy Revealed by MWISP -- the Galactic Longitude 45°--60°and 120°--130°

Molecular clouds (MCs) are cradles of star and planet formation, thereby playing an important role in the evolution of galaxies. Based on the unbiased Milky Way Imaging Scroll Painting (MWISP) survey data of $^{12}$CO, $^{13}$CO, and C$^{18}$O (J=1--0) line emission in two regions toward the inner and outer Galaxy, i.e. the G50 ($44.75°\le l \le 60.25°$) and G120 ($119.75°\le l \le 130.25°$) regions, the distribution of molecular gas is studied. Both regions have Galactic latitudes of $|b| \le 5.25°$. A catalog containing 24724 MCs is constructed from the data. In our proximity, several molecular structures with large angular scales and small velocity dispersions are discovered, resembling curtains of mist. Beyond the nearby molecular gas, a clear aggregation of MCs along coherent structures in the Galactic plane is visible, sketching spiral arm structures. Nevertheless, the aggregation of MCs is also detected in the inter-arm region between the Perseus and Outer arms in the G50 region. The Galactic molecular disk in this inter-arm region is found to be thinner than that in the adjacent spiral arm region. In addition, the thickness of the Galactic molecular disk examined here is found to be correlated with the warp of it, indicating their homologous origins. The molecular disk has a typical thickness of ~220 pc in the inner Galaxy. Moreover, the dispersion of the MC systemic velocity decreases with increasing galactocentric radius, resulting in lower kinematic distance uncertainties at larger radii. However, the Perseus arm segment in the G120 region exhibits a relatively large cloud-to-cloud velocity dispersion and split components in its MC velocity distribution.

astro-ph.GA↗

arXiv2025

MMESGBench: Pioneering Multimodal Understanding and Complex Reasoning Benchmark for ESG Tasks

Environmental, Social, and Governance (ESG) reports are essential for evaluating sustainability practices, ensuring regulatory compliance, and promoting financial transparency. However, these documents are often lengthy, structurally diverse, and multimodal, comprising dense text, structured tables, complex figures, and layout-dependent semantics. Existing AI systems often struggle to perform reliable document-level reasoning in such settings, and no dedicated benchmark currently exists in ESG domain. To fill the gap, we introduce \textbf{MMESGBench}, a first-of-its-kind benchmark dataset targeted to evaluate multimodal understanding and complex reasoning across structurally diverse and multi-source ESG documents. This dataset is constructed via a human-AI collaborative, multi-stage pipeline. First, a multimodal LLM generates candidate question-answer (QA) pairs by jointly interpreting rich textual, tabular, and visual information from layout-aware document pages. Second, an LLM verifies the semantic accuracy, completeness, and reasoning complexity of each QA pair. This automated process is followed by an expert-in-the-loop validation, where domain specialists validate and calibrate QA pairs to ensure quality, relevance, and diversity. MMESGBench comprises 933 validated QA pairs derived from 45 ESG documents, spanning across seven distinct document types and three major ESG source categories. Questions are categorized as single-page, cross-page, or unanswerable, with each accompanied by fine-grained multimodal evidence. Initial experiments validate that multimodal and retrieval-augmented models substantially outperform text-only baselines, particularly on visually grounded and cross-page tasks. MMESGBench is publicly available as an open-source dataset at https://github.com/Zhanglei1103/MMESGBench.

cs.MM↗

arXiv2025

Natively Trainable Sparse Attention for Hierarchical Point Cloud Datasets

Unlocking the potential of transformers on datasets of large physical systems depends on overcoming the quadratic scaling of the attention mechanism. This work explores combining the Erwin architecture with the Native Sparse Attention (NSA) mechanism to improve the efficiency and receptive field of transformer models for large-scale physical systems, addressing the challenge of quadratic attention complexity. We adapt the NSA mechanism for non-sequential data, implement the Erwin NSA model, and evaluate it on three datasets from the physical sciences -- cosmology simulations, molecular dynamics, and air pressure modeling -- achieving performance that matches or exceeds that of the original Erwin model. Additionally, we reproduce the experimental results from the Erwin paper to validate their implementation.

cs.LG↗

arXiv2025

HERMES: A Unified Self-Driving World Model for Simultaneous 3D Scene Understanding and Generation

Driving World Models (DWMs) have become essential for autonomous driving by enabling future scene prediction. However, existing DWMs are limited to scene generation and fail to incorporate scene understanding, which involves interpreting and reasoning about the driving environment. In this paper, we present a unified Driving World Model named HERMES. We seamlessly integrate 3D scene understanding and future scene evolution (generation) through a unified framework in driving scenarios. Specifically, HERMES leverages a Bird's-Eye View (BEV) representation to consolidate multi-view spatial information while preserving geometric relationships and interactions. We also introduce world queries, which incorporate world knowledge into BEV features via causal attention in the Large Language Model, enabling contextual enrichment for understanding and generation tasks. We conduct comprehensive studies on nuScenes and OmniDrive-nuScenes datasets to validate the effectiveness of our method. HERMES achieves state-of-the-art performance, reducing generation error by 32.4% and improving understanding metrics such as CIDEr by 8.0%. The model and code will be publicly released at https://github.com/LMD0311/HERMES.

cs.CV↗

arXiv2025

Semantic Item Graph Enhancement for Multimodal Recommendation

Multimodal recommendation systems have attracted increasing attention for their improved performance by leveraging items' multimodal information. Prior methods often build modality-specific item-item semantic graphs from raw modality features and use them as supplementary structures alongside the user-item interaction graph to enhance user preference learning. However, these semantic graphs suffer from semantic deficiencies, including (1) insufficient modeling of collaborative signals among items and (2) structural distortions introduced by noise in raw modality features, ultimately compromising performance. To address these issues, we first extract collaborative signals from the interaction graph and infuse them into each modality-specific item semantic graph to enhance semantic modeling. Then, we design a modulus-based personalized embedding perturbation mechanism that injects perturbations with modulus-guided personalized intensity into embeddings to generate contrastive views. This enables the model to learn noise-robust representations through contrastive learning, thereby reducing the effect of structural noise in semantic graphs. Besides, we propose a dual representation alignment mechanism that first aligns multiple semantic representations via a designed Anchor-based InfoNCE loss using behavior representations as anchors, and then aligns behavior representations with the fused semantics by standard InfoNCE, to ensure representation consistency. Extensive experiments on four benchmark datasets validate the effectiveness of our framework.

cs.IR↗

arXiv2025

CM$^3$: Calibrating Multimodal Recommendation

Alignment and uniformity are fundamental principles within the domain of contrastive learning. In recommender systems, prior work has established that optimizing the Bayesian Personalized Ranking (BPR) loss contributes to the objectives of alignment and uniformity. Specifically, alignment aims to draw together the representations of interacting users and items, while uniformity mandates a uniform distribution of user and item embeddings across a unit hypersphere. This study revisits the alignment and uniformity properties within the context of multimodal recommender systems, revealing a proclivity among extant models to prioritize uniformity to the detriment of alignment. Our hypothesis challenges the conventional assumption of equitable item treatment through a uniformity loss, proposing a more nuanced approach wherein items with similar multimodal attributes converge toward proximal representations within the hyperspheric manifold. Specifically, we leverage the inherent similarity between items' multimodal data to calibrate their uniformity distribution, thereby inducing a more pronounced repulsive force between dissimilar entities within the embedding space. A theoretical analysis elucidates the relationship between this calibrated uniformity loss and the conventional uniformity function. Moreover, to enhance the fusion of multimodal features, we introduce a Spherical Bézier method designed to integrate an arbitrary number of modalities while ensuring that the resulting fused features are constrained to the same hyperspherical manifold. Empirical evaluations conducted on five real-world datasets substantiate the superiority of our approach over competing baselines. We also shown that the proposed methods can achieve up to a 5.4% increase in NDCG@20 performance via the integration of MLLM-extracted features. Source code is available at: https://github.com/enoche/CM3.

cs.IR↗

arXiv2025

The Missing Covariate Indicator Method is Nearly Valid Almost Always

Background: Although the missing covariate indicator method (MCIM) has been shown to be biased under extreme conditions, the degree and determinants of bias have not been formally assessed. We derived the formula for the relative bias in the MCIM and systematically investigated conditions under which bias arises. We found that the extent of bias is independent of both the disease rate and the exposure-outcome association, but it is a function of 5 parameters: exposure and covariate prevalences, covariate missingness proportion, and associations of covariate with exposure and outcome. The MCIM was unbiased when the missing covariate is a risk factor for the outcome but not a confounder. The average median relative bias was zero across each of the parameters over a wide range of values considered. Our simulation study demonstrated that the mean and median of relative bias of MCIM was comparable to that of the no missingness method, which used the full sample with complete information for all variables, as long as the missingness of covariate is independent of the outcome. When missingness was no greater than 50%, less than 5% of the scenarios considered had relative bias greater than 10%. In several analyses of the Harvard cohort studies, the MCIM produced materially the same results as the multiple imputation method. In conclusion, the MCIM is nearly valid almost always in settings typically encountered in epidemiology and its continued use is recommended, unless the covariate is missing in an extreme proportion or acts as a strong confounder.

stat.AP↗

arXiv2025

Parameter-Efficient Fine-Tuning in Spectral Domain for Point Cloud Learning

Recently, leveraging pre-training techniques to enhance point cloud models has become a prominent research topic. However, existing approaches typically require full fine-tuning of pre-trained models to achieve satisfactory performance on downstream tasks, which is storage-intensive and computationally demanding. To address this issue, we propose a novel Parameter-Efficient Fine-Tuning (PEFT) method for point cloud, called \textbf{PointGST} (\textbf{Point} cloud \textbf{G}raph \textbf{S}pectral \textbf{T}uning). PointGST freezes the pre-trained model and introduces a lightweight, trainable Point Cloud Spectral Adapter (PCSA) for fine-tuning parameters in the spectral domain. The core idea is built on two observations: 1) The inner tokens from frozen models might present confusion in the spatial domain; 2) Task-specific intrinsic information is important for transferring the general knowledge to the downstream task. Specifically, PointGST transfers the point tokens from the spatial domain to the spectral domain, effectively de-correlating confusion among tokens by using orthogonal components for separation. Moreover, the generated spectral basis involves intrinsic information about the downstream point clouds, enabling more targeted tuning. As a result, PointGST facilitates the efficient transfer of general knowledge to downstream tasks while significantly reducing training costs. Extensive experiments on challenging point cloud datasets across various tasks demonstrate that PointGST not only outperforms its fully fine-tuning counterpart but also significantly reduces trainable parameters, making it a promising solution for efficient point cloud learning. The code will be made available at https://github.com/jerryfeng2003/PointGST

cs.CV↗

arXiv2025

SLICEMATE: Accurate and Scalable Static Program Slicing via LLM-Powered Agents

Static program slicing, which extracts the executable portions of a program that affect the values at a specific location, supports many software analysis tasks such as debugging and security auditing. However, traditional slicing tools rely on computationally expensive reachability analysis over dependency graphs, which struggle to scale to large programs and often fail to handle code with incomplete syntax. Recently emerged learning-based methods, while more robust to such cases, still fall short of achieving comparable performance to traditional methods on well-formed code. In this work, we propose SliceMate, a novel static program slicing solution powered by Large Language Model (LLM) agents. It bypasses the need for explicit dependency graph construction and achieving superior slicing accuracy. Concretely, SliceMate integrates three specialized agents: (1) a synthesis agent that produces candidate slices by incrementally expanding the scan scope across functions and files guided by LLM-inferred dependencies; (2) a verification agent that performs conciseness and completeness checks of the candidate slices, detecting missing or irrelevant statements; and (3) a refinement agent that repairs the slices with minimal edits in accordance with the verification results. These agents are orchestrated by a control module that ensures timely convergence and outputs high-quality slices without manual intervention. For rigorous evaluation, we construct a new and high-quality benchmark, SliceBench, comprising 2,200 manually annotated Java and Python programs, with program lengths ranging from 5 to 8,577 lines, significantly larger than those in existing slicing benchmarks. Experimental results show that SliceMate greatly outperforms both traditional and learning-based slicing tools.

cs.SE↗

arXiv2025

Phase transitions in voting simulated by an intelligent Ising model

Voting is an important social activity for expressing public opinions. By conceptually considering a group of voting agents to be intelligent matter, the impact of real-time information on voting results is quantitatively studied by an intelligent Ising model, which is formed by adding nonlinear instantaneous feedback of the overall magnetization to the conventional Ising model. In the new model, the interaction strength becomes a variable depending on the total magnetization rather than a constant, which mimics the scenario that the decision of an individual during vote influenced by the dynamically changing polling result during the election process. Our analytical derivations along with Mote Carlo simulations reveal that, with a positive feedback, the intelligent Ising model exhibits phase transitions at any finite temperatures, a feature lacked in the conventional one-dimensional Ising model. In all dimensions, by varying the feedback strength, the system changes from going through a second-order phase transition to going through a first-order phase transition with increasing temperature, and the two types of phase transitions are connected by a tricritical point. This study on the one hand demonstrates that the intelligent matter with a nonlinear adaptive interaction can exhibit qualitatively different phase behaviors from conventional matter, and on the other hand shows that, during voting, even unbiased feedback may possibly induce spontaneous symmetry breaking, leading to a biased outcome where one side of the vote becomes favored.

cond-mat.stat-mech↗

arXiv2025

miMamba: EEG-based Emotion Recognition with Multi-scale Inverted Mamba Models

EEG-based emotion recognition holds significant potential in the field of brain-computer interfaces. A key challenge lies in extracting discriminative spatiotemporal features from electroencephalogram (EEG) signals. Existing studies often rely on domain-specific time-frequency features and analyze temporal dependencies and spatial characteristics separately, neglecting the interaction between local-global relationships and spatiotemporal dynamics. To address this, we propose a novel network called Multi-Scale Inverted Mamba (MS-iMamba), which consists of Multi-Scale Temporal Blocks (MSTB) and Temporal-Spatial Fusion Blocks (TSFB). Specifically, MSTBs are designed to capture both local details and global temporal dependencies across different scale subsequences. The TSFBs, implemented with an inverted Mamba structure, focus on the interaction between dynamic temporal dependencies and spatial characteristics. The primary advantage of MS-iMamba lies in its ability to leverage reconstructed multi-scale EEG sequences, exploiting the interaction between temporal and spatial features without the need for domain-specific time-frequency feature extraction. Experimental results on the DEAP, DREAMER, and SEED datasets demonstrate that MS-iMamba achieves classification accuracies of 94.86%, 94.94%, and 91.36%, respectively, using only four-channel EEG signals, outperforming state-of-the-art methods.

cs.HC↗

arXiv2025

Power-efficient ultra-broadband soliton microcombs in resonantly-coupled microresonators

The drive to miniaturize optical frequency combs for practical deployment has spotlighted microresonator solitons as a promising chip-scale candidate. However, these soliton microcombs could be very power-hungry when their span increases, especially with fine comb spacings. As a result, realizing an octave-spanning comb at microwave repetition rates for direct optical-microwave linkage is considered not possible for photonic integration due to the high power requirements. Here, we introduce the concept of resonant-coupling to soliton microcombs to reduce pump consumption significantly. Compared to conventional waveguide-coupled designs, we demonstrate (i) a threefold increase in spectral span for high-power combs and (ii) up to a tenfold reduction in repetition frequency for octave-spanning operation. This configuration is compatible with laser integration and yields reliable, turnkey soliton generation. By eliminating the long-standing pump-power bottleneck, microcombs will soon become readily available for portable optical clocks, massively parallel data links, and field-deployable spectrometers.

physics.optics↗

arXiv2025

First Return, Entropy-Eliciting Explore

Reinforcement Learning from Verifiable Rewards (RLVR) improves the reasoning abilities of Large Language Models (LLMs) but it struggles with unstable exploration. We propose FR3E (First Return, Entropy-Eliciting Explore), a structured exploration framework that identifies high-uncertainty decision points in reasoning trajectories and performs targeted rollouts to construct semantically grounded intermediate feedback. Our method provides targeted guidance without relying on dense supervision. Empirical results on mathematical reasoning benchmarks(AIME24) show that FR3E promotes more stable training, produces longer and more coherent responses, and increases the proportion of fully correct trajectories. These results highlight the framework's effectiveness in improving LLM reasoning through more robust and structured exploration.

cs.AI↗

arXiv2025

One Size Does Not Fit All: Investigating Efficacy of Perplexity in Detecting LLM-Generated Code

Large language model-generated code (LLMgCode) has become increasingly common in software development. So far LLMgCode has more quality issues than human-authored code (HaCode). It is common for LLMgCode to mix with HaCode in a code change, while the change is signed by only human developers, without being carefully examined. Many automated methods have been proposed to detect LLMgCode from HaCode, in which the perplexity-based method (PERPLEXITY for short) is the state-of-the-art method. However, the efficacy evaluation of PERPLEXITY has focused on detection accuracy. Yet it is unclear whether PERPLEXITY is good enough in a wider range of realistic evaluation settings. To this end, we carry out a family of experiments to compare PERPLEXITY against feature- and pre-training-based methods from three perspectives: detection accuracy, detection speed, and generalization capability. The experimental results show that PERPLEXITY has the best generalization capability while having limited detection accuracy and detection speed. Based on that, we discuss the strengths and limitations of PERPLEXITY, e.g., PERPLEXITY is unsuitable for high-level programming languages. Finally, we provide recommendations to improve PERPLEXITY and apply it in practice. As the first large-scale investigation on detecting LLMgCode from HaCode, this article provides a wide range of findings for future improvement.

cs.SE↗

arXiv2025

Less is Enough: Training-Free Video Diffusion Acceleration via Runtime-Adaptive Caching

Video generation models have demonstrated remarkable performance, yet their broader adoption remains constrained by slow inference speeds and substantial computational costs, primarily due to the iterative nature of the denoising process. Addressing this bottleneck is essential for democratizing advanced video synthesis technologies and enabling their integration into real-world applications. This work proposes EasyCache, a training-free acceleration framework for video diffusion models. EasyCache introduces a lightweight, runtime-adaptive caching mechanism that dynamically reuses previously computed transformation vectors, avoiding redundant computations during inference. Unlike prior approaches, EasyCache requires no offline profiling, pre-computation, or extensive parameter tuning. We conduct comprehensive studies on various large-scale video generation models, including OpenSora, Wan2.1, and HunyuanVideo. Our method achieves leading acceleration performance, reducing inference time by up to 2.1-3.3$\times$ compared to the original baselines while maintaining high visual fidelity with a significant up to 36% PSNR improvement compared to the previous SOTA method. This improvement makes our EasyCache a efficient and highly accessible solution for high-quality video generation in both research and practical applications. The code is available at https://github.com/H-EmbodVis/EasyCache.

cs.CV↗

arXiv2025

SOFARI: High-Dimensional Manifold-Based Inference

Multi-task learning is a widely used technique for harnessing information from various tasks. Recently, the sparse orthogonal factor regression (SOFAR) framework, based on the sparse singular value decomposition (SVD) within the coefficient matrix, was introduced for interpretable multi-task learning, enabling the discovery of meaningful latent feature-response association networks across different layers. However, conducting precise inference on the latent factor matrices has remained challenging due to the orthogonality constraints inherited from the sparse SVD constraints. In this paper, we suggest a novel approach called the high-dimensional manifold-based SOFAR inference (SOFARI), drawing on the Neyman near-orthogonality inference while incorporating the Stiefel manifold structure imposed by the SVD constraints. By leveraging the underlying Stiefel manifold structure that is crucial to enabling inference, SOFARI provides easy-to-use bias-corrected estimators for both latent left factor vectors and singular values, for which we show to enjoy the asymptotic mean-zero normal distributions with estimable variances. We introduce two SOFARI variants to handle strongly and weakly orthogonal latent factors, where the latter covers a broader range of applications. We illustrate the effectiveness of SOFARI and justify our theoretical results through simulation examples and a real data application in economic forecasting.

stat.ME↗

arXiv2025

Integrated optomechanical ultrasonic sensors with nano-Pascal-level sensitivity

Ultrasonic sensors are widely used for object detection and localization in underwater and biological settings. The operational range and spatial resolution are inherently limited by sensor sensitivity, in which conventional piezoelectric transducers have been overwhelmed by advanced photonic sensors. Here, we demonstrate an optomechanical ultrasonic sensor integrated into a photonic platform, which comprises a suspended SiO2 membrane embedded with a high-Q Si3N4 microring resonator. By exploiting simultaneous optical and mechanical resonances, the sensor achieves a record low noise-equivalent pressure (NEP) of 218 nPa/Hz^1/2 at 289 kHz in air and 9.6 nPa/Hz^1/2 at 52 kHz in water. We demonstrate its versatility through photoacoustic gas spectroscopy in air and underwater ultrasound imaging, achieving a minimum detectable C2H2 concentration of 2.9 ppm (integration time 1 s) and an imaging resolution of 1.89 mm, respectively. Our work represents a significant advancement in compact CMOS-compatible ultrasound sensing, unlocking new possibilities in biomedical imaging, environmental monitoring, industrial testing, and underwater communications.

physics.optics↗

arXiv2025

UniTac-NV: A Unified Tactile Representation For Non-Vision-Based Tactile Sensors

Generalizable algorithms for tactile sensing remain underexplored, primarily due to the diversity of sensor modalities. Recently, many methods for cross-sensor transfer between optical (vision-based) tactile sensors have been investigated, yet little work focus on non-optical tactile sensors. To address this gap, we propose an encoder-decoder architecture to unify tactile data across non-vision-based sensors. By leveraging sensor-specific encoders, the framework creates a latent space that is sensor-agnostic, enabling cross-sensor data transfer with low errors and direct use in downstream applications. We leverage this network to unify tactile data from two commercial tactile sensors: the Xela uSkin uSPa 46 and the Contactile PapillArray. Both were mounted on a UR5e robotic arm, performing force-controlled pressing sequences against distinct object shapes (circular, square, and hexagonal prisms) and two materials (rigid PLA and flexible TPU). Another more complex unseen object was also included to investigate the model's generalization capabilities. We show that alignment in latent space can be implicitly learned from joint autoencoder training with matching contacts collected via different sensors. We further demonstrate the practical utility of our approach through contact geometry estimation, where downstream models trained on one sensor's latent representation can be directly applied to another without retraining.

cs.RO↗