arXiv subjects

Yuliang Chen

Publications and source records attributed to Yuliang Chen.

18 recordsLinked to original sources

arXiv2026

BALMS: Benchmarking Agentic LLMs for Longitudinal Mental Health Sensing

Mental health assessment relies on episodic self-report scales, which convert subjective states such as stress into numerical scores but provide only sparse snapshots of wellbeing. Wearable devices offer longitudinal behavioral and physiological signals for continuous, low-burden monitoring. Recent LLM-driven personal-health agents enable natural language queries over wearable signals, but mainly handle short-term, retrieval-based lookups (e.g., highest step count over a week). They do not evaluate whether agents can reason over long-term signals to predict wellbeing scores paired with evidence-grounded rationales. To address this gap, we introduce BALMS, the first systematic benchmark of LLM-based agentic systems for longitudinal mental health sensing. BALMS spans 3 real-world longitudinal datasets, 2 task families (closed-form wellbeing-score prediction and rationale generation auto-graded by an LLM-as-Judge), 3 agentic paradigms evaluated across 5 open- and closed-source LLM backbones. We find that zero-shot agents rarely outperform a simple mean baseline, except with stronger backbones or compact, semantically meaningful features. Chain-of-thought prompting improves reasoning-oriented backbones, but does not guarantee temporal grounding or numerical correctness. Together with more analysis on efficiency and temporal scaling, BALMS highlights the need for longitudinal mental health agents that selectively retrieve history, ground temporal evidence, and reason over interpretable behavioral features.

cs.CL↗

arXiv2026

Sedentary Behavior Classification for Wearable Sensors with a CNN-BiLSTM Model

Accurate detection of sedentary behavior is important for studying health risks related to prolonged sitting, but posture-based classification remains challenging with wearable sensors, especially at the wrist. We study whether a deep learning model trained on hip-worn accelerometer data can transfer to wrist-worn accelerometer data for sitting versus non-sitting classification. We use CHAP, a CNN-BiLSTM model originally developed for hip accelerometers, and evaluate its zero-shot performance on wrist data as well as its adaptation through finetuning with varying amounts of labeled wrist data. Experiments are conducted on the iWatch dataset with ground-truth posture labels derived from wearable cameras. The hip-trained model performs strongly on hip data without retraining, but accuracy drops on wrist data due to sensor placement shift. Finetuning CHAP provides consistent advantages over transformer models trained from scratch. These findings suggest that hip-based pretraining provides a useful starting point for wrist deployment, while highlighting the need for wrist-specific adaptation to handle higher signal variability.

cs.LG↗

arXiv2026

Toward World Modeling of Physiological Signals with Chaos-Theoretic Balancing and Latent Dynamics

Physiological time series signals reflect complex, multi-scale dynamical processes of the human body. Existing modeling studies focus on static tasks such as classification, event forecasting, or short-horizon next step prediction, while long-horizon signal-level forecasting and predictive nature of physiological signals remain underexplored. We introduce NormWear-2, a world model that encodes both multivariate physiological signals and clinical intervention variables into a shared latent space and models their joint temporal evolution as a dynamical system. Our approach combines inference from prior pre-trained knowledge (intuition) with instant non-parametric latent state transition adaptation (insight), enabling coherent forecasting across multiple temporal scales, conditioned on heterogeneous clinical interventions. During the pretraining phase, we find that chaos-theoretic balancing of dynamical regime diversity yields more robust representations, with a smaller balanced corpus outperforming one twice its size and capturing bifurcation regimes. We evaluate the world model performance across diverse real-world physiological datasets spanning heterogeneous temporal resolutions and intervention regimes, covering daily life, point-of-care, and clinical settings, including fitness planning, hemodialysis, diabetes management, and surgical monitoring. These evaluation datasets comprise records from 8,026 subjects, spanning study durations from 3.2 hours for high-resolution signal data to 2.3 years for longitudinal clinical biomarker tracking. NormWear-2 achieves the best overall forecasting performance across time, frequency, and latent representation domains, with significant improvements over state-of-the-art time series foundation models, while maintaining competitive downstream representation quality, providing a step toward general-purpose world models for physiological signals.

cs.LG↗

arXiv2026

Benchmarking Safety Risks of Knowledge-Intensive Reasoning under Malicious Knowledge Editing

Large language models (LLMs) increasingly rely on knowledge editing to support knowledge-intensive reasoning, but this flexibility also introduces critical safety risks: adversaries can inject malicious or misleading knowledge that corrupts downstream reasoning and leads to harmful outcomes. Existing knowledge editing benchmarks primarily focus on editing efficacy and lack a unified framework for systematically evaluating the safety implications of edited knowledge on reasoning behavior. To address this gap, we present EditRisk-Bench, a benchmark for systematically evaluating safety risks of knowledge-intensive reasoning under malicious knowledge editing. Unlike prior benchmarks that mainly emphasize edit success, generalization, and locality, EditRisk-Bench focuses on how injected knowledge affects downstream reasoning behavior and reliability. It integrates diverse malicious scenarios, including misinformation, bias, and safety violations, together with multi-level knowledge-intensive reasoning tasks and representative editing strategies within a unified evaluation framework measuring attack effectiveness, reasoning correctness, and side effects. Extensive experiments on both open-source and closed-source LLMs show that malicious knowledge editing can reliably induce incorrect or unsafe reasoning while largely preserving general capabilities, making such risks difficult to detect. We further identify several key factors influencing these risks, including edit scale, knowledge characteristics, and reasoning complexity. EditRisk-Bench provides an extensible testbed for understanding and mitigating safety risks in knowledge editing for LLMs.

cs.AI↗

arXiv2026

DSIPA: Detecting LLM-Generated Texts via Sentiment-Invariant Patterns Divergence Analysis

The rapid advancement of large language models (LLMs) presents new security challenges, particularly in detecting machine-generated text used for misinformation, impersonation, and content forgery. Most existing detection approaches struggle with robustness against adversarial perturbation, paraphrasing attacks, and domain shifts, often requiring restrictive access to model parameters or large labeled datasets. To address this, we propose DSIPA, a novel training-free framework that detects LLM-generated content by quantifying sentiment distributional stability under controlled stylistic variation. It is based on the observation that LLMs typically exhibit more emotionally consistent outputs, while human-written texts display greater affective variation. Our framework operates in a zero-shot, black-box manner, leveraging two unsupervised metrics, sentiment distribution consistency and sentiment distribution preservation, to capture these intrinsic behavioral asymmetries without the need for parameter updates or probability access. Extensive experiments are conducted on state-of-the-art proprietary and open-source models, including GPT-5.2, Gemini-1.5-pro, Claude-3, and LLaMa-3.3. Evaluations on five domains, such as news articles, programming code, student essays, academic papers, and community comments, demonstrate that DSIPA improves F1 detection scores by up to 49.89% over baseline methods. The framework exhibits superior generalizability across domains and strong resilience to adversarial conditions, providing a robust and interpretable behavioral signal for secure content identification in the evolving LLM landscape.

cs.CL↗

arXiv2026

CoopGuard: Stateful Cooperative Agents Safeguarding LLMs Against Evolving Multi-Round Attacks

As Large Language Models (LLMs) are increasingly deployed in complex applications, their vulnerability to adversarial attacks raises urgent safety concerns, especially those evolving over multi-round interactions. Existing defenses are largely reactive and struggle to adapt as adversaries refine strategies across rounds. In this work, we propose CoopGuard , a stateful multi-round LLM defense framework based on cooperative agents that maintains and updates an internal defense state to counter evolving attacks. It employs three specialized agents (Deferring Agent, Tempting Agent, and Forensic Agent) for complementary round-level strategies, coordinated by System Agent, which conditions decisions on the evolving defense state (interaction history) and orchestrates agents over time. To evaluate evolving threats, we introduce the EMRA benchmark with 5,200 adversarial samples across 8 attack types, simulating progressively LLM multi-round attacks. Experiments show that CoopGuard reduces attack success rate by 78.9% over state-of-the-art defenses, while improving deceptive rate by 186% and reducing attack efficiency by 167.9%, offering a more comprehensive assessment of multi-round defense. These results demonstrate that CoopGuard provides robust protection for LLMs in multi-round adversarial scenarios.

cs.CR↗

arXiv2026

Learning Transferable Sensor Models via Language-Informed Pretraining

Modern sensing systems generate large volumes of unlabeled multivariate time-series data. This abundance of unlabeled data makes self-supervised learning (SSL) a natural approach for learning transferable representations. However, most existing approaches are optimized for reconstruction or forecasting objectives and often fail to capture the semantic structure required for downstream classification and reasoning tasks. While recent sensor-language alignment methods improve semantic generalization through captioning and zero-shot transfer, they are limited to fixed sensor configurations, such as predefined channel sets, signal lengths, or temporal resolutions, which hinders cross-domain applicability. To address these gaps, we introduce \textbf{SLIP} (\textbf{S}ensor \textbf{L}anguage-\textbf{I}nformed \textbf{P}retraining), an open-source framework for learning language-aligned representations that generalize across diverse sensor setups. SLIP integrates contrastive alignment with sensor-conditioned captioning, facilitating both discriminative understanding and generative reasoning. By repurposing a pretrained decoder-only language model via cross-attention and introducing an elegant, flexible patch-embedder, SLIP supports different temporal resolutions and variable-length input at inference time without additional retraining. Across 11 datasets, SLIP demonstrates superior performance in zero-shot transfer, signal captioning, and question answering. It achieves a 77.14% average linear-probing accuracy, a 5.93% relative improvement over strong baselines, and reaches 64.83% accuracy in sensor-based question answering.

cs.AI↗

arXiv2026

Multi-Modal Style Transfer-based Prompt Tuning for Efficient Federated Domain Generalization

Federated Domain Generalization (FDG) aims to collaboratively train a global model across distributed clients that can generalize well on unseen domains. However, existing FDG methods typically struggle with cross-client data heterogeneity and incur significant communication and computation overhead. To address these challenges, this paper presents a new FDG framework, dubbed FaST-PT, which facilitates local feature augmentation and efficient unseen domain adaptation in a distributed manner. First, we propose a lightweight Multi-Modal Style Transfer (MST) method to transform image embedding under text supervision, which could expand the training data distribution and mitigate domain shift. We then design a dual-prompt module that decomposes the prompt into global and domain prompts. Specifically, global prompts capture general knowledge from augmented embedding across clients, while domain prompts capture domain-specific knowledge from local data. Besides, Domain-aware Prompt Generation (DPG) is introduced to adaptively generate suitable prompts for each sample, which facilitates unseen domain adaptation through knowledge fusion. Extensive experiments on four cross-domain benchmark datasets, e.g., PACS and DomainNet, demonstrate the superior performance of FaST-PT over SOTA FDG methods such as FedDG-GA and DiPrompt. Ablation studies further validate the effectiveness and efficiency of FaST-PT.

cs.DC↗

arXiv2025

MoCA: Multi-modal Cross-masked Autoencoder for Time Series in Digital Health

Wearable devices enable continuous multi-modal physiological and behavioral monitoring, yet analysis of these data streams faces fundamental challenges including the lack of gold-standard labels and incomplete sensor data. While self-supervised learning approaches have shown promise for addressing these issues, existing multi-modal extensions present opportunities to better leverage the rich temporal and cross-modal correlations inherent in simultaneously recorded wearable sensor data. We propose the Multi-modal Cross-masked Autoencoder (MoCA), a self-supervised learning framework that combines transformer architecture with masked autoencoder (MAE) methodology, using a principled cross-modality masking scheme that explicitly leverages correlation structures between sensor modalities. MoCA demonstrates strong performance boosts across reconstruction and downstream classification tasks on diverse benchmark datasets. We further establish theoretical guarantees by establishing a fundamental connection between multi-modal MAE loss and kernelized canonical correlation analysis through a Reproducing Kernel Hilbert Space framework, providing principled guidance for correlation-aware masking strategy design. Our approach offers a novel solution for leveraging unlabeled multi-modal wearable data while handling missing modalities, with broad applications across digital health domains.

stat.ML↗

arXiv2024

Toward Foundation Model for Multivariate Wearable Sensing of Physiological Signals

Time-series foundation models excel at tasks like forecasting across diverse data types by leveraging informative waveform representations. Wearable sensing data, however, pose unique challenges due to their variability in patterns and frequency bands, especially for healthcare-related outcomes. The main obstacle lies in crafting generalizable representations that adapt efficiently across heterogeneous sensing configurations and applications. To address this, we propose NormWear, the first multi-modal and ubiquitous foundation model designed to extract generalized and informative representations from wearable sensing data. Specifically, we design a channel-aware attention mechanism with a shared special liaison [CLS] token to detect signal patterns in both intra-sensor and inter-sensors. This helps the model to extract more meaningful information considering both time series themselves and the relationships between input sensors. This helps the model to be widely compatible with various sensors settings. NormWear is pretrained on a diverse set of physiological signals, including PPG, ECG, EEG, GSR, and IMU, from various public datasets. Our model shows exceptional generalizability across 11 public wearable sensing datasets, spanning 18 applications in mental health, body state inference, vital sign estimation, and disease risk evaluation. It consistently outperforms competitive baselines under zero-shot, partial-shot, and full-shot settings, indicating broad applicability in real-world health applications.

cs.LG↗

arXiv2021

Manipulating the anisotropic phase separation in strained VO2 epitaxial films by nanoscale ion-implantation

Manipulating the strain induced poly-domains and phase transition in correlated oxide material are important for high performance devices fabrication. Though the electronic transport in the strained oxide film at macroscopic scales can be directly measured, the anisotropic electronic state and the controllable phase separation cross the insulator-to-metal transition within nanoscale size are still elusive. Here, we selected VO2 crystal film as a prototypical oxide and achieved the manipulation of anisotropy electronic phase separation via injecting He+ nanobeam into VO2 film at room temperature. In addition, this nanoscale phase separation was directly visualized by infrared near-field imaging measurements, showing the pronounced and unique cR-axis dependent anisotropy on VO2 surface. Our results offered new insights towards understanding the anisotropic nanoscale phase separation in strained metal oxide films.

cond-mat.mtrl-sci↗

arXiv2021

Spatially-resolved insulator-metal transition for rewritable optical gratings

Doping is an effective way to tune the property of metal oxides1-5, for achieving functional oxide electronics6-8. Previously we developed a controllable hydrogen doping technology at ambient conditions by use of electron-proton synergistic doping strategy, which enables one to get rid of high-temperature/pressure treatments required by traditional technologies9. Here, based on this facile doping route, we achieve a visual and reversible insulator-metal transition (MIT) for tungsten trioxide (WO3) film. Its outstanding spatial selection is comparable to standard UV lithography, which shows the potential of becoming a viable way for rewritable WO3 grating device fabrication. Furthermore, the period of the obtained WO3 structural grating can also be easily changed for requirement by doping area selection. This advanced doping technology opens up alternative approaches for developing not only optical devices, but also rewritable ions devices and integrated circuits for various oxide electronics.

physics.app-ph↗

arXiv2020

Visualizing an adjustable WO3/p-GaN heterojunction

The p-n junctions based on typical semiconductors are the elementary units for the modern electronic devices and chip industry. While the rectification property of those p-n junction is usually fixed once the unit is fabricated. Here, we proposed an adjustable n-WO3/p-GaN heterojunction with controllable electronic properties. For the prepared n-WO3/p-GaN heterojunction, it is almost transparent and shows typical p-n junction rectification. While if gradually doping some hydrogen atoms into WO3 layer by a facile electron-proton synergistic route, the heterojunction can be turned dynamically from the typical p-n junction (n-WO3/p-GaN) to standard Schottky contact (HxWO3/p-GaN) step by step. More importantly, this evolution can be directly visualized by eyesight due to the pronounced electrochromic characteristic of WO3 layer. By connecting two HxWO3/p-GaN heterojunctions, the controllable bi-functional rectification can be achieved. In addition, the HxWO3/p-GaN heterojunction can recovered to the original p-n jucntion just by annealing at ambient, demonstrating the heterojunction is controllable and reusable. The current study will open up tremendous opportunities for dynamic electronic devices in the future.

physics.app-ph↗

arXiv2019

Uniaxial Strain-Induced Anisotropic Charge Transfer in Contact Electrification

Even though contact electrification has been studied for a long time, the mechanism of charge transfer still remains elusive. Most of previous reports only focus on the driving force of charge transfer. However, to better understand the mechanism of charge transfer, we believe contact itself for supplying transferring path that charges taking also need to be understood. Here, we focus on the in-plane symmetry of contact in contact electrification by utilizing a uniaxial strain to change the material's isotropic nature to anisotropic. A clear anisotropic charge transfer is observed by contacting axially stretched rubber films at different rotational angles, which could arise from fluctuation of contacting area in microscale. A universal ellipse model is also proposed for qualitatively describing the anisotropy of contact regardless of the specific driving force of contact electrification.

physics.app-ph↗

arXiv2019

Electron-proton Co-doping Induced Metal-insulator Transition in VO2 Film via Surface Self-assembled Ascorbic Acid Molecules

Charge doping is an effective way to induce metal-insulate transition (MIT) in correlated materials for many important utilizations, which is however practically limited by problem of low stability. In this study, we have achieved pronounced phase modulation and stabilized the metallic state of monoclinic vanadium dioxide (VO2) at room temperature, via a novel electron-proton co-doping mechanism driven by surface absorption of self-assembled L-ascorbic acid (AA) molecules. The ionized AA- species in solution donate effective electrons to the adsorbed VO2 surface, which then electrostatically attract surrounding protons to penetrate, and eventually results in stable hydrogen-doped metallic VO2. The variations of phase and electronic structures as well as the electron occupancy of V-3d/O-2p hybrid orbitals were examined by synchrotron characterizations and first-principle theoretical simulations, which explain the formation of stable metallic state. Importantly, the adsorbed molecules protect hydrogen dopants from escaping out of lattice and thereby stabilize the metallic phase for VO2. Such an electron-proton co-doping mechanism driven by suitable molecules absorption would open a new door for engineering properties of correlated oxide materials.

physics.app-ph↗

arXiv2019

Gate-Controlled VO2 Phase Transition for High-Performance Smart Window

VO2 material is promising for developing energy-saving "smart window", owing to its thermochromic property induced by metal-insulator transition (MIT). However, its practical application is greatly limited by the relatively high critical transition temperature (~68oC), low luminous transmittance (<60%) and poor solar energy regulation ability (<15%). Here we developed a reversible and non-volatile electric-field control on the MIT of monoclinic VO2 film. With a solid electrolyte layer assisted gating treatment, we modulated the insertion/extraction of hydrogens into/from VO2 lattice at room temperature, causing tri-state phase transitions accompanied with controllable transmission adjustment. The dramatic increase of visible/infrared transmittance during the phase transition from the metallic (lightly H-doping) to insulating (heavily H-doping) phase leads to an increased solar energy regulation ability up to 26.5%, while keep 70.8% visible-luminous transmittance. These results beat all previous records and even exceeded the theoretical limit for traditional VO2 smart window, removing intrinsic disadvantages of VO2 for energy-saving utilizations. Our findings not only demonstrated an electric-field controlled phase modulation strategy, but also open the door for high-performance VO2-based smart window applications.

physics.app-ph↗

arXiv2017

Hydrogenating VO2 with protons in acid solution

Hydrogenation is an effective way to tune material property1-5. Traditional techniques for doping hydrogen atoms into solid materials are very costly due to the need for noble metal catalysis and high-temperature/pressure annealing treatment or even high energy proton implantation in vacuum condition5-8. Acid solution contains plenty of freely-wandering protons, but it is difficult to act as a proton source for doping, since the protons always cause corrosions by destroying solid lattices before residing into them. Here we achieve a facile way to hydrogenate monoclinic vanadium dioxide (VO2) with protons in acid solution by attaching suitable metal to it. Considering the Schottky contact at the metal/VO2 interface, electrons flow from metal to VO2 due to workfunction difference and simultaneously attract free protons in acid solution to penetrate, forming the hydrogens dopants inside VO2 lattice. This metal-acid treatment constitutes an electron-proton co-doping strategy, which not only protects the VO2 lattice from corrosion, but also causes pronounced insulator-to-metal transitions. In addition, the metal-acid induced hydrogen doping behavior shows a ripple effect, and it can spread contagiously up to wafer-size area (>2 inch) even triggered by a tiny metal particle attachment (~1.0mm). This will stimulate a new way of simple and cost-effective atomic doping technique for some other oxide materials.

cond-mat.mtrl-sci↗

arXiv2017

Consecutive Insulator-Metal-Insulator Phase Transitions of Vanadium Dioxide by Hydrogen Doping

We report modulation of a reversible phase transition in VO2 films by hydrogen doping. A metallic phase and a new insulating phase are successively observed at room temperature as the doping concentration increases. It is suggested that the polarized charges from doped hydrogens play an important role. These charges gradually occupy V3d-O2p hybridized orbitals and consequently modulate the filling of the VO2 crystal conduction band-edge states, which eventually evolve into new valence band-edge states. This demonstrates the exceptional sensitivity of VO2 electronic properties to electron concentration and orbital occupancy, providing key information for the phase transition mechanism.

cond-mat.str-el↗