arXiv subjects

Yang Xu

Publications and source records attributed to Yang Xu.

At least 181 records · Page 10Linked to original sources

arXiv2025

Event-aware analysis of cross-city visitor flows using large language models and social media data

Public events, such as music concerts and fireworks displays, can cause irregular surges in cross-city travel demand, leading to potential overcrowding, travel delays, and public safety concerns. To better anticipate and accommodate such demand surges, it is essential to estimate cross-city visitor flows with awareness of public events. Although prior studies typically focused on the effects of a single mega event or disruptions around a single venue, this study introduces a generalizable framework to analyze visitor flows under diverse and concurrent events. We propose to leverage large language models (LLMs) to extract event features from multi-source online information and massive user-generated content on social media platforms. Specifically, social media popularity metrics are designed to capture the effects of online promotion and word-of-mouth in attracting visitors. An event-aware machine learning model is then adopted to uncover the specific impacts of different event features and ultimately predict visitor flows for upcoming events. Using Hong Kong as a case study, the framework is applied to predict daily flows of mainland Chinese visitors arriving at the city, achieving a testing R-squared of over 85%. We further investigate the heterogeneous event impacts on visitor numbers across different event types and major travel modes. Both promotional popularity and word-of-mouth popularity are found to be associated with increased visitor flows, but the specific effects vary by the event type. This association is more pronounced among visitors arriving by metro and high-speed rail, while it has less effect on air travelers. The findings can facilitate coordinated measures across government agencies and guide specialized transport policies, such as shuttle transit services to event venues, and comprehensive on-site traffic management strategies.

cs.SI↗

arXiv2025

Cross-region Model Training with Communication-Computation Overlapping and Delay Compensation

Training large language models (LLMs) requires massive computational resources, often necessitating the aggregation of geographically distributed data centers (\ie, cross-region training). However, the high communication latency in wide-area networks severely degrades the efficiency of traditional distributed training. While methods like DiLoCo reduce communication frequency, they suffer from blocking synchronization. Streaming DiLoCo alleviates this issue via communication-computation overlapping but introduces update staleness and model inconsistency due to delayed global updates and partial synchronization. These factors impair convergence, especially when aggressive overlap is needed to mask high latency. We propose CoCoDC, a novel distributed training framework with communication-computation overlapping and delay compensation, to explicitly tackle these challenges. Within the CoCoDC framework, we specifically develop a novel Delay Compensation strategy based on Taylor expansion to effectively mitigate the staleness and an Adaptive Transmission strategy that dynamically schedules model fragment synchronization to optimize bandwidth usage and accelerate convergence. Extensive experiments highlight the superior performance of CoCoDC over both DiLoCo and Streaming DiLoCo regarding final accuracy and training speed. Specifically, CoCoDC reduces the training steps needed to reach a comparable perplexity by up to 21.0% compared to Streaming DiLoCo. Our work provides an effective solution for scalable and efficient cross-region LLM training.

cs.DC↗

arXiv2025

Visual moral inference and communication

Humans can make moral inferences from multiple sources of input. In contrast, automated moral inference in artificial intelligence typically relies on language models with textual input. However, morality is conveyed through modalities beyond language. We present a computational framework that supports moral inference from natural images, demonstrated in two related tasks: 1) inferring human moral judgment toward visual images and 2) analyzing patterns in moral content communicated via images from public news. We find that models based on text alone cannot capture the fine-grained human moral judgment toward visual stimuli, but language-vision fusion models offer better precision in visual moral inference. Furthermore, applications of our framework to news data reveal implicit biases in news categories and geopolitical discussions. Our work creates avenues for automating visual moral inference and discovering patterns of visual moral communication in public media.

cs.CV↗

arXiv2025

Distributed Formation Shape Control of Identity-less Robot Swarms

Different from most of the formation strategies where robots require unique labels to identify topological neighbors to satisfy the predefined shape constraints, we here study the problem of identity-less distributed shape formation in homogeneous swarms, which is rarely studied in the literature. The absence of identities creates a unique challenge: how to design appropriate target formations and local behaviors that are suitable for identity-less formation shape control. To address this challenge, we propose the following novel results. First, to avoid using unique identities, we propose a dynamic formation description method and solve the formation consensus of robots in a locally distributed manner. Second, to handle identity-less distributed formations, we propose a fully distributed control law for homogeneous swarms based on locally sensed information. While the existing methods are applicable to simple cases where the target formation is stationary, ours can tackle more general maneuvering formations such as translation, rotation, or even shape deformation. Both numerical simulation and flight experiment are presented to verify the effectiveness and robustness of our proposed formation strategy.

cs.RO↗

arXiv2025

Hybrid neural network method of a multilayer perceptron and autoencoder for the α-particle preformation factor in α-decay theory

The preformation factor quantifies the probability of α particles preforming on the surface of the parent nucleus in decay theory and is closely related to the study of α clustering structure. In this work, a multilayer perceptron and autoencoder (MLP + AE) hybrid neural network method is introduced to extract preformation factors within the generalized liquid drop model and experimental data. A K-fold cross validation method is also adopted. The accuracy of the preformation factor calculated by this improved neural network is comparable to the results of the empirical formula. MLP + AE can effectively capture the linear relationship between the logarithm of the preformation factor and the square root of the ratio of the decay energy, further verifying that Geiger-Nuttall law can deal with preformation factor. The extracted preformation probability of isotope and isotone chains show different trends near the magic number, and in addition, an odd-even staggering effect appears. This means that the preformation factors are affected by closed shells and unpaired nucleons. Therefore the preformation factors can provide nuclear structure information. Furthermore, for 41 new nuclides, the half-lives introduced with the preformation factors reproduce the experimental values as expected. Finally, the preformation factors and α-decay half-lives of Z = 119 and 120 superheavy nuclei are predicted.

nucl-th↗

arXiv2025

Pay More Attention to the Robustness of Prompt for Instruction Data Mining

Instruction tuning has emerged as a paramount method for tailoring the behaviors of LLMs. Recent work has unveiled the potential for LLMs to achieve high performance through fine-tuning with a limited quantity of high-quality instruction data. Building upon this approach, we further explore the impact of prompt's robustness on the selection of high-quality instruction data. This paper proposes a pioneering framework of high-quality online instruction data mining for instruction tuning, focusing on the impact of prompt's robustness on the data mining process. Our notable innovation, is to generate the adversarial instruction data by conducting the attack for the prompt of online instruction data. Then, we introduce an Adversarial Instruction-Following Difficulty metric to measure how much help the adversarial instruction data can provide to the generation of the corresponding response. Apart from it, we propose a novel Adversarial Instruction Output Embedding Consistency approach to select high-quality online instruction data. We conduct extensive experiments on two benchmark datasets to assess the performance. The experimental results serve to underscore the effectiveness of our proposed two methods. Moreover, the results underscore the critical practical significance of considering prompt's robustness.

cs.AI↗

arXiv2025

High-fidelity spatial information transfer through dynamic scattering media by an epsilon-near-zero time-gate

Transparent conducting oxides (TCO) such as indium-tin-oxide (ITO) exhibit strong optical nonlinearity in the frequency range where their permittivities are near zero. We leverage this nonlinear optical response to realize a sub-picosecond time-gate based on upconversion (or sum-) four-wave mixing (FWM) between two ultrashort pulses centered at the epsilon-near-zero (ENZ) wavelength in a sub-micron-thick ITO film. The time-gate removes the effect of both static and dynamic scattering on the signal pulse by retaining only the ballistic photons of the pulse, that is, the photons that are not scattered. Thus, the spatial information encoded in either the intensity or the phase of the signal pulse can be preserved and transmitted with high fidelity through scattering media. Furthermore, in the presence of time-varying scattering, our time-gate can reduce the resulting scintillation by two orders of magnitude. In contrast to traditional bulk nonlinear materials, time gating by sum-FWM in a sub-wavelength-thick ENZ film can produce a scattering-free upconverted signal at a visible wavelength without sacrificing spatial resolution, which is usually limited by the phase-matching condition. Our proof-of-principle experiment can have implications for potential applications such as \textit{in vivo} diagnostic imaging and free-space optical communication.

physics.optics↗

arXiv2025

Resource-Efficient Federated Fine-Tuning Large Language Models for Heterogeneous Data

Fine-tuning large language models (LLMs) via federated learning, i.e., FedLLM, has been proposed to adapt LLMs for various downstream applications in a privacy-preserving way. To reduce the fine-tuning costs on resource-constrained devices, FedLoRA is proposed to fine-tune only a small subset of model parameters by integrating low-rank adaptation (LoRA) into FedLLM. However, apart from resource constraints, there is still another critical challenge, i.e., data heterogeneity, severely hindering the implementation of FedLoRA in practical applications. Herein, inspired by the previous group-based federated learning paradigm, we propose a hierarchical FedLoRA framework, termed HierFedLoRA, to address these challenges. Specifically, HierFedLoRA partitions all devices into multiple near-IID groups and adjusts the intra-group aggregation frequency for each group to eliminate the negative effects of non-IID data. Meanwhile, to reduce the computation and communication cost, HierFedLoRA dynamically assigns diverse and suitable fine-tuning depth (i.e., the number of continuous fine-tuning layers from the output) for each group. HierFedLoRA explores jointly optimizing aggregation frequency and depth upon their coupled relationship to better enhance the performance of FedLoRA. Extensive experiments are conducted on a physical platform with 80 commercial devices. The results show that HierFedLoRA improves the final model accuracy by 1.6% to 4.2%, speeding up the fine-tuning process by at least 2.1$\times$, compared to the strong baselines.

cs.LG↗

arXiv2025

Ultralow-pressure mechanical-motion switching of ferroelectric polarization

Ferroelectric polarization switching, achieved by mechanical forces, enables the storage of stress information in ferroelectrics, and holds promise for human-interfacing applications. The prevailing mechanical approach is locally induced flexoelectricity with large strain gradients. However, this approach usually requires huge mechanical pressures, which greatly impedes device applications. Here, we report an approach of using triboelectric effect to mechanically, reversibly switch ferroelectric polarization across α-In2Se3 ferroelectric memristors. Through contact electrification and electrostatic induction effects, triboelectric units are used to sensitively detect mechanical forces and generate electrical voltage pulses to trigger α-In2Se3 resistance switching. We realize multilevel resistance states under different mechanical forces, by which a neuromorphic stress system is demonstrated. Strikingly, we achieve the reversal of α-In2Se3 ferroelectric polarization with a record-low mechanical pressure of ~ 10 kPa, and even with tactile touches. Our work provides a fundamental but pragmatic strategy for creating mechanical-tactile ferroelectric memory devices.

cond-mat.mtrl-sci↗

arXiv2025

A Novel Hat-Shaped Device-Cloud Collaborative Inference Framework for Large Language Models

Recent advancements in large language models (LLMs) have catalyzed a substantial surge in demand for LLM services. While traditional cloud-based LLM services satisfy high-accuracy requirements, they fall short in meeting critical demands for low delay and enhanced privacy. To address these limitations, we propose HAT, a novel device-cloud collaborative inference framework that leverages the complementary strengths of U-shaped inference and speculative decoding. HAT partitions the LLM into three submodels, and the input and output submodels, stacked with a lightweight adapter network, are deployed as a small language model (SLM) on each end device. Meanwhile, the middle submodel, encompassing the majority of the LLM's decoder layers, is hosted in the cloud to perform speculative decoding with on-device SLMs. During inference, HAT exchanges hidden states (rather than raw tokens) of input or draft tokens between devices and the cloud, thereby incurring substantial communication delays. Besides, processing hidden states of long prompts will exacerbate computation delays in the cloud, further compromising inference efficiency. To improve efficiency, we introduce a prompt chunking mechanism that segments long prompts into shorter chunks, enabling parallel transmission and processing. Furthermore, HAT is implemented to dynamically determine optimal chunk sizes for devices handling long prompts, thereby improving overall inference speed. Extensive experiments are conducted on a physical testbed comprising 30 NVIDIA Jetson devices and a server with 8 NVIDIA A6000 GPUs. Experimental results demonstrate that HAT achieves promising performance improvements, reducing TTFT by 41% to 54% and TBT by 41% to 77% compared to the baselines.

cs.LG↗

arXiv2025

Multistate Density Functional Theory for Local and Charge-Transfer Tripdoublet States from Triplet-Free Radical Interactions

The interaction between excited states of a closed-shell chromophore and a nearby free radical species gives rise to spin-coupled doublet states, namely singdoublet and tripdoublet, as well as a quartet state. This coupling facilitates transitions that are otherwise spin-forbidden, thereby enhancing intersystem crossing and influencing luminescence and non-radiative decay pathways. In this chapter, we explore these interactions using multistate density functional theory (MSDFT). By employing a minimal active space (MAS) comprising just ten determinant configurations, MSDFT effectively captures local and charge-transfer excitations with inclusion of correlation effects. MSDFT extends the Hohenberg-Kohn density functional theory from the ground state to encompass all electronic states, underscoring the potential for developing computationally efficient methods to study excited states. Numerical results demonstrate that MSDFT accurately reproduces both qualitative trends and quantitative excited-state energies, in accord with previous studies using extended multistate complete-active-space second-order perturbation theory (XMS-CASPT2). The work explores energy changes along a reaction path from the D_0/D_1 minimum energy crossing intersection to the D_2/D_3 crossing in the exciplex formed by 10-methylphenothiazine and a dicarboximide electron acceptor linked to the stable free radical 2,2,6,6-tetramethylpiperidin-1-oxyl (TEMPO).

physics.chem-ph↗

arXiv2025

OpenCoder: The Open Cookbook for Top-Tier Code Large Language Models

Large language models (LLMs) for code have become indispensable in various domains, including code generation, reasoning tasks and agent systems. While open-access code LLMs are increasingly approaching the performance levels of proprietary models, high-quality code LLMs suitable for rigorous scientific investigation, particularly those with reproducible data processing pipelines and transparent training protocols, remain limited. The scarcity is due to various challenges, including resource constraints, ethical considerations, and the competitive advantages of keeping models advanced. To address the gap, we introduce OpenCoder, a top-tier code LLM that not only achieves performance comparable to leading models but also serves as an "open cookbook" for the research community. Unlike most prior efforts, we release not only model weights and inference code, but also the reproducible training data, complete data processing pipeline, rigorous experimental ablation results, and detailed training protocols for open scientific research. Through this comprehensive release, we identify the key ingredients for building a top-tier code LLM: (1) code optimized heuristic rules for data cleaning and methods for data deduplication, (2) recall of text corpus related to code and (3) high-quality synthetic data in both annealing and supervised fine-tuning stages. By offering this level of openness, we aim to broaden access to all aspects of a top-tier code LLM, with OpenCoder serving as both a powerful model and an open foundation to accelerate research, and enable reproducible advancements in code AI.

cs.CL↗

arXiv2025

The Implicit Bias of Heterogeneity towards Invariance: A Study of Multi-Environment Matrix Sensing

Models are expected to engage in invariance learning, which involves distinguishing the core relations that remain consistent across varying environments to ensure the predictions are safe, robust and fair. While existing works consider specific algorithms to realize invariance learning, we show that model has the potential to learn invariance through standard training procedures. In other words, this paper studies the implicit bias of Stochastic Gradient Descent (SGD) over heterogeneous data and shows that the implicit bias drives the model learning towards an invariant solution. We call the phenomenon the implicit invariance learning. Specifically, we theoretically investigate the multi-environment low-rank matrix sensing problem where in each environment, the signal comprises (i) a lower-rank invariant part shared across all environments; and (ii) a significantly varying environment-dependent spurious component. The key insight is, through simply employing the large step size large-batch SGD sequentially in each environment without any explicit regularization, the oscillation caused by heterogeneity can provably prevent model learning spurious signals. The model reaches the invariant solution after certain iterations. In contrast, model learned using pooled SGD over all data would simultaneously learn both the invariant and spurious signals. Overall, we unveil another implicit bias that is a result of the symbiosis between the heterogeneity of data and modern algorithms, which is, to the best of our knowledge, first in the literature.

cs.LG↗

arXiv2025

Stealthy Jailbreak Attacks on Large Language Models via Benign Data Mirroring

Large language model (LLM) safety is a critical issue, with numerous studies employing red team testing to enhance model security. Among these, jailbreak methods explore potential vulnerabilities by crafting malicious prompts that induce model outputs contrary to safety alignments. Existing black-box jailbreak methods often rely on model feedback, repeatedly submitting queries with detectable malicious instructions during the attack search process. Although these approaches are effective, the attacks may be intercepted by content moderators during the search process. We propose an improved transfer attack method that guides malicious prompt construction by locally training a mirror model of the target black-box model through benign data distillation. This method offers enhanced stealth, as it does not involve submitting identifiable malicious instructions to the target model during the search phase. Our approach achieved a maximum attack success rate of 92%, or a balanced value of 80% with an average of 1.5 detectable jailbreak queries per sample against GPT-3.5 Turbo on a subset of AdvBench. These results underscore the need for more robust defense mechanisms.

cs.CL↗

arXiv2025

Programmable Electric Tweezers

The interaction mechanism between a single microscopic object like a cell, a particle, a molecule, or an atom and its interacting electromagnetic field is fundamental in single-object manipulation such as optical trap and magnetic trap. Function-on-demand, single-object manipulation relies on a high degree of freedom control of electromagnetic field at localized scales, which remains challenging. Here we propose a manipulation concept: programmable single-object manipulation, based on programming the electromagnetic field in a multi-bit electrode system. This concept is materialized on a Programmable Electric Tweezer (PET) with four individually addressed electrodes, marking a transition from function-fixed single-object manipulation to function-programmable single-object manipulation. By programming the localized electric field, our PET can provide various manipulation functions for achieving precise trapping, movement and rotation of multiscale single microscopic objects, including single proteins, nucleic acids, microparticles and bacteria. Implementing these functions, we are able not only to manipulate the object of interest on demand but also quantitatively measure the charge to mass ratio of a single microparticle via the Paul trap and the electrical properties of an individual bacterial cell by the rotation analysis. Finally, with superposed single-particle trapping and rotation, we demonstrate the spontaneous relaxation of DNA supercoiling and observe an unexpected pause phenomenon in the relaxation process, highlighting the versatility and the potential of PET in uncovering stochastic biophysical phenomena at the single-molecule level.

physics.bio-ph↗

arXiv2025

Pressure Tuning of Layer-hybridized Excitons in Trilayer WSe2

We demonstrate dynamic pressure tuning (0-6.6 GPa) of layer-hybridized excitons in AB-stacked trilayer WSe$_2$ via diamond-anvil-cell-integrated reflectance spectroscopy. Pressure-controlled interlayer coupling manifests in enhanced energy-level anti-crossings and oscillator strength redistribution, with Stark shift analysis revealing a characteristic dipole moment reduction of 11%. Notably, the hybridization strength between the intra- and interlayer excitons triples from $\sim$10 meV to above $\sim$30 meV, exhibiting a near-linear scaling of 3.5$\pm$0.2 meV/GPa. Spectral density simulations resolve four distinct components, i.e., intralayer ground/excited and interlayer ground/excited excitons, with their relative weights transitioning from one component dominant to strongly hybridized at higher pressures. Our findings highlight the potential for controlling excitonic properties and engineering novel optoelectronic devices through interlayer compression.

cond-mat.mtrl-sci↗

arXiv2025

A Review of Causal Decision Making

To make effective decisions, it is important to have a thorough understanding of the causal relationships among actions, environments, and outcomes. This review aims to surface three crucial aspects of decision-making through a causal lens: 1) the discovery of causal relationships through causal structure learning, 2) understanding the impacts of these relationships through causal effect learning, and 3) applying the knowledge gained from the first two aspects to support decision making via causal policy learning. Moreover, we identify challenges that hinder the broader utilization of causal decision-making and discuss recent advances in overcoming these challenges. Finally, we provide future research directions to address these challenges and to further enhance the implementation of causal decision-making in practice, with real-world applications illustrated based on the proposed causal decision-making. We aim to offer a comprehensive methodology and practical implementation framework by consolidating various methods in this area into a Python-based collection. URL: https://causaldm.github.io/Causal-Decision-Making.

stat.ML↗

arXiv2025

Electrothermal manipulation of current-induced phase transitions in ferrimagnetic Mn$_3$Si$_2$Te$_6$

Phase transitions driven by external stimuli are central to condensed matter physics, providing critical insights into symmetry breaking and emergent phenomena. Recently, ferrimagnetic (FiM) Mn$_3$Si$_2$Te$_6$ has attracted considerable attention for its magnetic-field-induced insulator-metal transitions (IMTs) and unconventional current-driven phase transitions, yet the role of applied currents in the magnetic phase remains poorly understood. Here, by combining local magnetization probes and time-resolved transport measurements, we uncover an electrothermal origin for the current-induced first-order-like phase transitions, characterized by abrupt voltage jumps and distinct magnetic domain evolution. Current-voltage (I-V) characteristics measured under triangular waveforms exhibit strong non-reciprocal and hysteretic behaviors, which are significantly suppressed at frequencies ~1000 Hz. Time-resolved studies using rectangular pulsed currents demonstrate that the resistance dynamics closely mirror the equilibrium resistance-temperature profile, directly implicating Joule heating as the driving mechanism. Furthermore, we reveal that the intrinsic I-V response adheres to Ohm's law, displaying linearity across various magnetic fields and temperatures. Our work advocates for a cautious approach in distinguishing between genuine current-induced nonequilibrium quantum states and thermal effects.

cond-mat.mtrl-sci↗