arXiv subjects

Jie Li

Publications and source records attributed to Jie Li.

At least 145 records · Page 8Linked to original sources

arXiv2026

Federated Cross-Modal Retrieval with Missing Modalities via Semantic Routing and Adapter Personalization

Federated cross-modal retrieval faces severe challenges from heterogeneous client data, particularly non-IID semantic distributions and missing modalities. Under such heterogeneity, a single global model is often insufficient to capture both shared cross-modal knowledge and client-specific characteristics. We propose RCSR, a personalization-friendly federated framework that integrates prototype anchoring, retrieval-centric semantic routing, and optional client-specific adapters. Built on a frozen CLIP backbone, RCSR leverages lightweight shared adapters for global knowledge transfer while supporting efficient local personalization. Prototype anchoring helps unimodal clients align with global cross-modal semantics, and a server-side semantic router adaptively assigns aggregation weights based on retrieval consistency to mitigate alignment drift during heterogeneous updates. Extensive experiments on MS-COCO, Flickr30K, and other benchmarks show that RCSR consistently improves global retrieval accuracy and training stability, while further enhancing client-level retrieval performance, especially for clients with incomplete modalities. Code is available at https://github.com/RezinChow/RCSR-Retrieval-Centric-Semantic-Routing.

cs.CV↗

arXiv2026

A Hierarchical MARL-Based Approach for Coordinated Retail P2P Trading and Wholesale Market Participation of DERs

The ongoing shift towards decentralization of the electric energy sector, driven by the growing electrification across end-use sectors, and widespread adoption of distributed energy resources (DERs), necessitates their active participation in the electricity markets to support grid operations. Furthermore, with bi-directional energy and communication flows becoming standard, intelligent, easy-to-deploy, resource-conservative demand-side participation is expected to play a critical role in securing power grid operational flexibility and market efficiency. This work proposes a market engagement framework that leverages a hierarchical multi-agent deep reinforcement learning (MARL) approach to enable individual prosumers to participate in peer-to-peer retail auctions and further aggregate these intelligent prosumers to facilitate effective DER participation in wholesale markets. Ultimately, a Stackelberg game is proposed to coordinate this hierarchical MARL-based DER market participation framework toward enhanced market performance.

cs.LG↗

arXiv2026

Interlayer hybridization enables superconductivity in bilayer nickelates

Ruddlesden-Popper nickelates offer a new route to high-temperature superconductivity beyond the cuprates and iron-pnictides. However, the electronic reorganization that enables superconductivity in bilayer nickelates remain unresolved, largely due to the difficulty of directly probing the superconducting phase. Here, we overcome this limitation by stabilizing superconducting (La,Pr)$_3$Ni$_2$O$_7$ thin films with a protective capping layer, thereby enabling direct spectroscopic access via X-ray absorption and resonant inelastic X-ray scattering. We resolve the evolution of in-plane and out-of-plane electronic states, spin and orbital excitations, and spin-density-waves across insulating, superconducting, and metallic regimes. Combining experimental results with theoretical analysis, we show that the in-plane $d_{x^2-y^2}$ states form an itinerant backbone, while superconductivity emerges only when coherent $d_{z^2}$-$p_z$-$d_{z^2}$ interlayer hybridization develops, accompanied by suppressed static spin order and strongly damped spin excitations. Oxygen stoichiometry and epitaxial strain both act on this interlayer channel, placing superconductivity within a narrow window of interlayer coherence and correlation strength. These findings identify the microscopic ingredients required for superconductivity in bilayer nickelates and provide a multiorbital picture of its emergence.

cond-mat.supr-con↗

arXiv2026

On the Optimization Landscape of Observer-based Dynamic Linear Quadratic Control

Understanding the optimization landscape of linear quadratic regulation (LQR) problems is fundamental to the design of efficient reinforcement learning solutions. Recent work has made significant progress in characterizing the landscape of static output-feedback control and linear quadratic Gaussian (LQG) control. For LQG, much of the analysis leverages the separation principle, which allows the controller and estimator to be designed independently. However, this simplification breaks down when the gradients with respect to the estimator and controller parameters are inherently coupled, leading to a more intricate analysis. This paper investigates the optimization landscape of observer-based dynamic output-feedback control of LQR problems. We derive the optimal observer-controller pair in settings where transient quadratic performance cannot be neglected. Our analysis reveals that, in general, the combination of the standard LQR controller and the observer that minimizes the trace of the accumulated estimation error covariance does not correspond to a stationary point of the overall closed-loop performance objective. Moreover, we derive a pair of discrete-time Sylvester equations with symmetric structure, both involving the same set of matrix elements, that characterize the stationary point of the observer-based dynamic LQR problem. These equations offer analytical insight into the structure of the optimality conditions and provide a foundation for developing numerical policy gradient methods aimed at learning complex controllers that rely on reconstructed state information.

eess.SY↗

arXiv2026

Hitem3D 2.0: Multi-View Guided Native 3D Texture Generation

Although recent advances have improved the quality of 3D texture generation, existing methods still struggle with incomplete texture coverage, cross-view inconsistency, and misalignment between geometry and texture. To address these limitations, we propose Hitem3D 2.0, a multi-view guided native 3D texture generation framework that enhances texture quality through the integration of 2D multi-view generation priors and native 3D texture representations. Hitem3D 2.0 comprises two key components: a multi-view synthesis framework and a native 3D texture generation model. The multi-view generation is built upon a pre-trained image editing backbone and incorporates plug-and-play modules that explicitly promote geometric alignment, cross-view consistency, and illumination uniformity, thereby enabling the synthesis of high-fidelity multi-view images. Conditioned on the generated views and 3D geometry, the native 3D texture generation model projects multi-view textures onto 3D surfaces while plausibly completing textures in unseen regions. Through the integration of multi-view consistency constraints with native 3D texture modeling, Hitem3D 2.0 significantly improves texture completeness, cross-view coherence, and geometric alignment. Experimental results demonstrate that Hitem3D 2.0 outperforms existing methods in terms of texture detail, fidelity, consistency, coherence, and alignment.

cs.CV↗

arXiv2026

Low-Confidence Gold: Refining Low-Confidence Samples for Efficient Instruction Tuning

The effectiveness of instruction fine-tuning for Large Language Models is fundamentally constrained by the quality and efficiency of training datasets. This work introduces Low-Confidence Gold (LCG), a novel filtering framework that employs centroid-based clustering and confidence-guided selection for identifying valuable instruction pairs. Through a semi-supervised approach using a lightweight classifier trained on representative samples, LCG curates high-quality subsets while preserving data diversity. Experimental evaluation demonstrates that models fine-tuned on LCG-filtered subsets of 6K samples achieve superior performance compared to existing methods, with substantial improvements on MT-bench and consistent gains across comprehensive evaluation metrics. The framework's efficacy while maintaining model performance establishes a promising direction for efficient instruction tuning.All open-source assets are publicly available at https://github.com/Lizruletheworld/Low-Confidence_Gold.

cs.CL↗

arXiv2026

MECO: A Multimodal Dataset for Emotion and Cognitive Understanding in Older Adults

While affective computing has advanced considerably, multimodal emotion prediction in aging populations remains underexplored, largely due to the scarcity of dedicated datasets. Existing multimodal benchmarks predominantly target young, cognitively healthy subjects, neglecting the influence of cognitive decline on emotional expression and physiological responses. To bridge this gap, we present MECO, a Multimodal dataset for Emotion and Cognitive understanding in Older adults. MECO includes 42 participants and provides approximately 38 hours of multimodal signals, yielding 30,592 synchronized samples. To maximize ecological validity, data collection followed standardized protocols within community-based settings. The modalities cover video, audio, electroencephalography (EEG), and electrocardiography (ECG). In addition, the dataset offers comprehensive annotations of emotional and cognitive states, including self-assessed valence, arousal, six basic emotions, and Mini-Mental State Examination cognitive scores. We further establish baseline benchmarks for both emotion and cognitive prediction. MECO serves as a foundational resource for multimodal modeling of affect and cognition in aging populations, facilitating downstream applications such as personalized emotion recognition and early detection of mild cognitive impairment (MCI) in real-world settings. The complete dataset and supplementary materials are available at https://maitrechen.github.io/meco-page/.

cs.HC↗

arXiv2026

Advanced Capacity Accreditation of Future Energy System Resources with Deep Uncertainties

The electric power sector has seen an increased penetration of renewable energy sources (RESs) that could strain the system reliability due to their inherent uncertainties in availability and controllability. Effective load carrying capability (ELCC) is widely used to quantify the reliability contributions of these RESs. However, existing ELCC methods can over- or under-estimate their contributions and often neglect or simplify other critical factors such as transmission constraints and evolving climate trends, leading to inaccurate capacity credit (CC) allocations and inefficient reliability procurement in capacity markets. To address these limitations, this paper proposes TRACED (TRansmission And Climate Enhanced Delta) -- an advanced capacity accreditation approach that integrates transmission constraints and climate-adjusted system conditions into a Delta ELCC evaluation. Case studies on a modified IEEE-118 bus system with high RES and energy storage penetrations demonstrate that TRACED produces portfolio-consistent CC allocations by capturing resource interactions and avoiding the double-counting of shared reliability benefits inherent in marginal ELCC, which may otherwise lead to under-procurement of reliability resources. Results further demonstrate that transmission congestion and evolving climate trends have mutual impacts on CC allocation, justifying their necessary integration into TRACED.

eess.SY↗

arXiv2026

Quantum control and signal enhancement exploiting the Stokes-anti-Stokes coherence

We present a theoretical framework for the coherent coupling between Stokes and anti-Stokes scattering processes, revealing interference phenomena inaccessible to either process alone. Within a dispersive-interaction model beyond the resolved-sideband limit, we show that classical driving and system linewidth coherently links the two channels, enabling phase-controlled interference. Destructive interference induces intrinsic asymmetry in dispersively coupled systems, enabling coherent control of quantum information storage and transfer, while constructive interference leads to exponential signal amplification and thus enhanced quantum detection. This work establishes a unified picture for understanding Stokes-anti-Stokes coherence as a fundamental mechanism underlying both quantum control and metrology. Furthermore, it suggests that these functionalities can be further enhanced by implementing Stokes-anti-Stokes arrays.

quant-ph↗

arXiv2026

NNQA: Neural-Native Quantum Arithmetic for End-to-End Polynomial Synthesis

Hybrid classical quantum learning is often bottlenecked by communication overhead and approximation error from generic variational ansatzes. In this study, we introduce Neural Native Quantum Arithmetic (NNQA), which compiles classically learned nonlinear representations into precise quantum arithmetic composed of native unitary blocks. Theoretically, we prove that the universal approximation of quantum polynomial arithmetic can be realized by transforming a classical neural network into a quantum circuit, with the resulting error arising solely from measurement shot noise, thereby extending classical operator-level estimation guarantees into the quantum regime. Empirical validation on IBM Quantum Heron3 and IonQ Forte processors shows performance limited primarily by device noise without variational fine tuning: we achieve over 99.5% accuracy for polynomials up to degree 35 and demonstrate scalability on IonQ hardware up to 36 qubits and circuit depths of 70, reaching a negligible RMSE of 0.005. Overall, NNQA establishes a new paradigm of synthesizing quantum arithmetic for native quantum computation.

quant-ph↗

arXiv2026

A Pipelined Collaborative Speculative Decoding Framework for Efficient Edge-Cloud LLM Inference

Recent advancements and widespread adoption of Large Language Models (LLMs) in both industry and academia have catalyzed significant demand for LLM serving. However, traditional cloud services incur high costs, while on-device inference alone faces challenges due to limited resources. Edge-cloud collaboration emerges as a key research direction to combine the strengths of both paradigms, yet efficiently utilizing limited network bandwidth while fully leveraging and balancing the computational capabilities of edge devices and the cloud remains an open problem. To address these challenges, we propose Pipelined Collaborative Speculative Decoding Framework (PicoSpec), a novel, general-purpose, and training-free speculative decoding framework for LLM edge-cloud collaborative inference. We design an asynchronous pipeline that resolves the mutual waiting problem inherent in vanilla speculative decoding within edge collaboration scenarios, which concurrently executes a Small Language Model (SLM) on the edge device and a LLM in the cloud. Meanwhile, to mitigate the significant communication latency caused by transmitting vocabulary distributions, we introduce separate rejection sampling with sparse compression, which completes the rejection sampling with only a one-time cost of transmitting the compressed vocabulary. Experimental results demonstrate that our solution outperforms baseline and existing methods, achieving up to 2.9 speedup.

cs.DC↗

arXiv2026

Xiaomi-Robotics-0: An Open-Sourced Vision-Language-Action Model with Real-Time Execution

In this report, we introduce Xiaomi-Robotics-0, an advanced vision-language-action (VLA) model optimized for high performance and fast and smooth real-time execution. The key to our method lies in a carefully designed training recipe and deployment strategy. Xiaomi-Robotics-0 is first pre-trained on large-scale cross-embodiment robot trajectories and vision-language data, endowing it with broad and generalizable action-generation capabilities while avoiding catastrophic forgetting of the visual-semantic knowledge of the underlying pre-trained VLM. During post-training, we propose several techniques for training the VLA model for asynchronous execution to address the inference latency during real-robot rollouts. During deployment, we carefully align the timesteps of consecutive predicted action chunks to ensure continuous and seamless real-time rollouts. We evaluate Xiaomi-Robotics-0 extensively in simulation benchmarks and on two challenging real-robot tasks that require precise and dexterous bimanual manipulation. Results show that our method achieves state-of-the-art performance across all simulation benchmarks. Moreover, Xiaomi-Robotics-0 can roll out fast and smoothly on real robots using a consumer-grade GPU, achieving high success rates and throughput on both real-robot tasks. To facilitate future research, code and model checkpoints are open-sourced at https://xiaomi-robotics-0.github.io

cs.RO↗

arXiv2026

GigaWorld-Policy: An Efficient Action-Centered World--Action Model

World-Action Models (WAM) initialized from pre-trained video generation backbones have demonstrated remarkable potential for robot policy learning. However, existing approaches face two critical bottlenecks that hinder performance and deployment. First, jointly reasoning over future visual dynamics and corresponding actions incurs substantial inference overhead. Second, joint modeling often entangles visual and motion representations, making motion prediction accuracy heavily dependent on the quality of future video forecasts. To address these issues, we introduce GigaWorld-Policy, an action-centered WAM that learns 2D pixel-action dynamics while enabling efficient action decoding, with optional video generation. Specifically, we formulate policy training into two coupled components: the model predicts future action sequences conditioned on the current observation, and simultaneously generates future videos conditioned on the predicted actions and the same observation. The policy is supervised by both action prediction and video generation, providing richer learning signals and encouraging physically plausible actions through visual-dynamics constraints. With a causal design that prevents future-video tokens from influencing action tokens, explicit future-video generation is optional at inference time, allowing faster action prediction during deployment. To support this paradigm, we curate a diverse, large-scale robot dataset to pre-train an action-centered video generation model, which is then adapted as the backbone for robot policy learning. Experimental results on real-world robotic platforms show that GigaWorld-Policy runs 9x faster than the leading WAM baseline, Motus, while improving task success rates by 7%. Moreover, compared with pi-0.5, GigaWorld-Policy improves performance by 95% on RoboTwin 2.0.

cs.CV↗

arXiv2026

Comment on: "Coherent perfect absorption: Zero reflection without linewidth suppression"

A recent paper, Phys. Rev. Research 8, 013261 (2026), claims that the polaromechanical normal-mode splitting (NMS) measured in Nat. Commun. 16, 5652 (2025) is not true based on their two results: $i$) there is no true splitting in the linear-scale spectrum; $ii$) the total or intrinsic decay rate of the cavity-magnon polariton, set by the imaginary part of the pole of the total output spectrum, remains unchanged under the coherent-perfect-absorption (CPA) condition. In this comment, we indicate that $i$) there is NMS in both the linear and logarithmic scales of our spectra in {\it a narrow frequency range} around the CPA frequency; $ii$) the total decay rate defined via the {\it pole} of the spectrum cannot characterize the vanishing {\it effective} decay rate at the CPA frequency (known as the monochromaticity of the CPA), and thus this parameter is irrelevant to the NMS measured in our experiment in {\it a narrow frequency range} around the CPA frequency. Consequently, their results above are either false or irrelevant, and thus cannot support their claim on the polaromechanical strong coupling measured in our experiment.

cond-mat.mes-hall↗

arXiv2026

FIND: A Simple yet Effective Baseline for Diffusion-Generated Image Detection

The remarkable realism of images generated by diffusion models poses critical detection challenges. Current methods utilize reconstruction error as a discriminative feature, exploiting the observation that real images exhibit higher reconstruction errors when processed through diffusion models. However, these approaches require costly reconstruction computations and depend on specific diffusion models, making their performance highly model-dependent. We identify a fundamental difference: real images are more difficult to fit with Gaussian distributions compared to synthetic ones. In this paper, we propose Forgery Identification via Noise Disturbance (FIND), a novel method that requires only a simple binary classifier. It eliminates reconstruction by directly targeting the core distributional difference between real and synthetic images. Our key operation is to add Gaussian noise to real images during training and label these noisy versions as synthetic. This step allows the classifier to focus on the statistical patterns that distinguish real from synthetic images. We theoretically prove that the noise-augmented real images resemble diffusion-generated images in their ease of Gaussian fitting. Furthermore, simply by adding noise, they still retain visual similarity to the original images, highlighting the most discriminative distribution-related features. The proposed FIND improves performance by 11.7% on the GenImage benchmark while running 126x faster than existing methods. By removing the need for auxiliary diffusion models and reconstruction, it offers a practical, efficient, and generalizable way to detect diffusion-generated content.

cs.CV↗

arXiv2026

GIAT: A Geologically-Informed Attention Transformer for Lithology Identification

Accurate lithology identification from well logs is crucial for subsurface resource evaluation. Although Transformer-based models excel at sequence modeling, their "black-box" nature and lack of geological guidance limit their performance and trustworthiness. To overcome these limitations, this letter proposes the Geologically-Informed Attention Transformer (GIAT), a novel framework that deeply fuses data-driven geological priors with the Transformer's attention mechanism. The core of GIAT is a new attention-biasing mechanism. We repurpose Category-Wise Sequence Correlation (CSC) filters to generate a geologically-informed relational matrix, which is injected into the self-attention calculation to explicitly guide the model toward geologically coherent patterns. On two challenging datasets, GIAT achieves state-of-the-art performance with an accuracy of up to 95.4%, significantly outperforming existing models. More importantly, GIAT demonstrates exceptional interpretation faithfulness under input perturbations and generates geologically coherent predictions. Our work presents a new paradigm for building more accurate, reliable, and interpretable deep learning models for geoscience applications.

cs.LG↗

arXiv2026

Nonlinear Conjugate Gradient Method for Multiobjective Optimization Problems of Interval-Valued Maps

In this article, we propose an algorithm for the nonlinear conjugate gradient method to find a Pareto critical point of unconstrained multiobjective interval optimization problems. In this algorithm, we use the Wolfe line search procedure to find the step length. After defining the standard Wolfe conditions and the strong Wolfe conditions, we prove that there exists an interval of the step length that satisfies the standard Wolfe conditions and the strong Wolfe conditions. Further, to study the convergence analysis of our proposed algorithm, we derive the result related to the Zoutendijk condition. In the convergence analysis, first, we prove the global convergence property of our proposed algorithm for a general conjugate gradient algorithmic parameter. Further, we consider four variants of the conjugate gradient algorithmic parameter, such as Fletcher-Reeves, conjugate descent, Dai-Yuan, and modified Dai-Yuan. For each variant of the algorithmic parameter, we prove the global convergence results of our proposed algorithm. Finally, we test our algorithm on some test problems and make a performance profile.

math.OC↗

arXiv2026

GhostEI-Bench: Do Mobile Agents Resilience to Environmental Injection in Dynamic On-Device Environments?

Vision-Language Models (VLMs) are increasingly deployed as autonomous agents to navigate mobile graphical user interfaces (GUIs). Operating in dynamic on-device ecosystems, which include notifications, pop-ups, and inter-app interactions, exposes them to a unique and underexplored threat vector: environmental injection. Unlike prompt-based attacks that manipulate textual instructions, environmental injection corrupts an agent's visual perception by inserting adversarial UI elements (for example, deceptive overlays or spoofed notifications) directly into the GUI. This bypasses textual safeguards and can derail execution, causing privacy leakage, financial loss, or irreversible device compromise. To systematically evaluate this threat, we introduce GhostEI-Bench, the first benchmark for assessing mobile agents under environmental injection attacks within dynamic, executable environments. Moving beyond static image-based assessments, GhostEI-Bench injects adversarial events into realistic application workflows inside fully operational Android emulators and evaluates performance across critical risk scenarios. We further propose a judge-LLM protocol that conducts fine-grained failure analysis by reviewing the agent's action trajectory alongside the corresponding screenshot sequence, pinpointing failure in perception, recognition, or reasoning. Comprehensive experiments on state-of-the-art agents reveal pronounced vulnerability to deceptive environmental cues: current models systematically fail to perceive and reason about manipulated UIs. GhostEI-Bench provides a framework for quantifying and mitigating this emerging threat, paving the way toward more robust and secure embodied agents.

cs.CR↗