arXiv subjects

Jie Liu

Publications and source records attributed to Jie Liu.

At least 235 records · Page 13Linked to original sources

arXiv2025

SurveyGen-I: Consistent Scientific Survey Generation with Evolving Plans and Memory-Guided Writing

Survey papers play a critical role in scientific communication by consolidating progress across a field. Recent advances in Large Language Models (LLMs) offer a promising solution by automating key steps in the survey-generation pipeline, such as retrieval, structuring, and summarization. However, existing LLM-based approaches often struggle with maintaining coherence across long, multi-section surveys and providing comprehensive citation coverage. To address these limitations, we introduce SurveyGen-I, an automatic survey generation framework that combines coarse-to-fine retrieval, adaptive planning, and memory-guided generation. SurveyGen-I first performs survey-level retrieval to construct the initial outline and writing plan, and then dynamically refines both during generation through a memory mechanism that stores previously written content and terminology, ensuring coherence across subsections. When the system detects insufficient context, it triggers fine-grained subsection-level retrieval. During generation, SurveyGen-I leverages this memory mechanism to maintain coherence across subsections. Experiments across four scientific domains demonstrate that SurveyGen-I consistently outperforms previous works in content quality, consistency, and citation coverage.

cs.CL↗

arXiv2025

EraserDiT: Fast Video Inpainting with Diffusion Transformer Model

Video object removal and inpainting are critical tasks in the fields of computer vision and multimedia processing, aimed at restoring missing or corrupted regions in video sequences. Traditional methods predominantly rely on flow-based propagation and spatio-temporal Transformers, but these approaches face limitations in effectively leveraging long-term temporal features and ensuring temporal consistency in the completion results, particularly when dealing with large masks. Consequently, performance on extensive masked areas remains suboptimal. To address these challenges, this paper introduces a novel video inpainting approach leveraging the Diffusion Transformer (DiT). DiT synergistically combines the advantages of diffusion models and transformer architectures to maintain long-term temporal consistency while ensuring high-quality inpainting results. We propose a Circular Position-Shift strategy to further enhance long-term temporal consistency during the inference stage. Additionally, the proposed method interactively removes specified objects, and generates corresponding prompts. In terms of processing speed, it takes only 65 seconds (testing on one NVIDIA H800 GPU) to complete a video with a resolution of $2160 \times 2100$ with 97 frames without any acceleration method. Experimental results indicate that the proposed method demonstrates superior performance in content fidelity, texture restoration, and temporal consistency. Project page:https://jieliu95.github.io/EraserDiT_demo/

cs.CV↗

arXiv2025

Lightweight Quad Bayer HybridEVS Demosaicing via State Space Augmented Cross-Attention

Event cameras like the Hybrid Event-based Vision Sensor (HybridEVS) camera capture brightness changes as asynchronous "events" instead of frames, offering advanced application on mobile photography. However, challenges arise from combining a Quad Bayer Color Filter Array (CFA) sensor with event pixels lacking color information, resulting in aliasing and artifacts on the demosaicing process before downstream application. Current methods struggle to address these issues, especially on resource-limited mobile devices. In response, we introduce \textbf{TSANet}, a lightweight \textbf{T}wo-stage network via \textbf{S}tate space augmented cross-\textbf{A}ttention, which can handle event pixels inpainting and demosaicing separately, leveraging the benefits of dividing complex tasks into manageable subtasks. Furthermore, we introduce a lightweight Cross-Swin State Block that uniquely utilizes positional prior for demosaicing and enhances global dependencies through the state space model with linear complexity. In summary, TSANet demonstrates excellent demosaicing performance on both simulated and real data of HybridEVS while maintaining a lightweight model, averaging better results than the previous state-of-the-art method DemosaicFormer across seven diverse datasets in both PSNR and SSIM, while respectively reducing parameter and computation costs by $1.86\times$ and $3.29\times$. Our approach presents new possibilities for efficient image demosaicing on mobile devices. Code is available in the supplementary materials.

cs.CV↗

arXiv2025

GLM-4.5: Agentic, Reasoning, and Coding (ARC) Foundation Models

We present GLM-4.5, an open-source Mixture-of-Experts (MoE) large language model with 355B total parameters and 32B activated parameters, featuring a hybrid reasoning method that supports both thinking and direct response modes. Through multi-stage training on 23T tokens and comprehensive post-training with expert model iteration and reinforcement learning, GLM-4.5 achieves strong performance across agentic, reasoning, and coding (ARC) tasks, scoring 70.1% on TAU-Bench, 91.0% on AIME 24, and 64.2% on SWE-bench Verified. With much fewer parameters than several competitors, GLM-4.5 ranks 3rd overall among all evaluated models and 2nd on agentic benchmarks. We release both GLM-4.5 (355B parameters) and a compact version, GLM-4.5-Air (106B parameters), to advance research in reasoning and agentic AI systems. Code, models, and more information are available at https://github.com/zai-org/GLM-4.5.

cs.CL↗

arXiv2025

Non-Hermitian superconducting diode effect

The study of non-reciprocal phenomena has long captivated interest in both Hermitian and non-Hermitian systems. The superconducting diode effect (SDE) is a non-reciprocal phenomenon characterized by unequal critical charge supercurrents flowing in opposite directions in Hermitian superconducting systems. In this study, we introduce an SDE driven by non-Hermiticity in a superconducting quantum interference device (SQUID) under an external magnetic flux, which we refer to as the non-Hermitian SDE. Non-Hermiticity is introduced by coupling one of the two Josephson junctions to a gapless electron reservoir, introducing phase decoherence. Remarkably, we find that an emergent non-Hermitian Fermi-Dirac distribution can give rise to SDE in the non-Hermitian SQUID. We analyze the behavior of the SDE under both direct current (dc) and alternating current (ac) biases, highlighting the appearance of direction-dependent critical currents and asymmetric Shapiro steps as hallmarks of the SDE. Our findings not only reveal an experimentally accessible mechanism for non-Hermitian SDE but also open new avenues for investigating non-reciprocal phenomena in non-Hermitian systems.

cond-mat.supr-con↗

arXiv2025

Quantum storage with flat bands

The realization of robust quantum storage devices relies on the ability to generate long-lived, spatially localized quantum states. In this work, we introduce a method for the targeted creation of compact excitations in flat-band lattices. By injecting in-plane radiation waves from the system's edge and applying a localized on-site potential at the desired storage position, we induce hybridization between compact localized states (CLSs) of the flat-band and resonant dispersive plane waves. This hybridization enables the formation of spatially compact, stable excitations suitable for quantum memory applications. We experimentally validate this mechanism using photonic waveguide arrays, focusing on two representative geometries: the diamond chain and the one-dimensional Lieb ladder. Our approach is broadly applicable to any platform supporting flat-band physics.

physics.optics↗

arXiv2025

DP-DGAD: A Generalist Dynamic Graph Anomaly Detector with Dynamic Prototypes

Dynamic graph anomaly detection (DGAD) is essential for identifying anomalies in evolving graphs across domains such as finance, traffic, and social networks. Recently, generalist graph anomaly detection (GAD) models have shown promising results. They are pretrained on multiple source datasets and generalize across domains. While effective on static graphs, they struggle to capture evolving anomalies in dynamic graphs. Moreover, the continuous emergence of new domains and the lack of labeled data further challenge generalist DGAD. Effective cross-domain DGAD requires both domain-specific and domain-agnostic anomalous patterns. Importantly, these patterns evolve temporally within and across domains. Building on these insights, we propose a DGAD model with Dynamic Prototypes (DP) to capture evolving domain-specific and domain-agnostic patterns. Firstly, DP-DGAD extracts dynamic prototypes, i.e., evolving representations of normal and anomalous patterns, from temporal ego-graphs and stores them in a memory buffer. The buffer is selectively updated to retain general, domain-agnostic patterns while incorporating new domain-specific ones. Then, an anomaly scorer compares incoming data with dynamic prototypes to flag both general and domain-specific anomalies. Finally, DP-DGAD employs confidence-based pseudo-labeling for effective self-supervised adaptation in target domains. Extensive experiments demonstrate state-of-the-art performance across ten real-world datasets from different domains.

cs.LG↗

arXiv2025

Interaction-Driven Altermagnetic Magnon Chiral Splitting

Nonrelativistic magnon chiral splitting in altermagnets has garnered significant recent attention. In this work, we demonstrate that nonlinear three-wave mixing -- where magnons split or coalesce -- extends this phenomenon into unprecedented relativistic regimes. Employing a bilayer antiferromagnet with Dzyaloshinskii-Moriya interactions, we identify three distinct classes of chiral splitting, each dictated by specific symmetries, such as $C_4T$, $σ_v T$, or their combination. This reveals a novel bosonic mechanism for symmetry-protected chiral splitting, capitalizing on the unique ability of magnons to violate particle-number conservation, a feature absent in low-energy fermionic systems. Our findings pave the way for engineering altermagnetic splitting, with potential applications in advanced magnonic devices and deeper insights into magnon dynamics in complex magnetic systems.

cond-mat.mtrl-sci↗

arXiv2025

AI PsyRoom: Artificial Intelligence Platform for Segmented Yearning and Reactive Outcome Optimization Method

Psychological counseling faces huge challenges due to the growing demand for mental health services and the shortage of trained professionals. Large language models (LLMs) have shown potential to assist psychological counseling, especially in empathy and emotional support. However, existing models lack a deep understanding of emotions and are unable to generate personalized treatment plans based on fine-grained emotions. To address these shortcomings, we present AI PsyRoom, a multi-agent simulation framework designed to enhance psychological counseling by generating empathetic and emotionally nuanced conversations. By leveraging fine-grained emotion classification and a multi-agent framework, we construct a multi-agent PsyRoom A for dialogue reconstruction, generating a high-quality dialogue dataset EmoPsy, which contains 35 sub-emotions, 423 specific emotion scenarios, and 12,350 dialogues. We also propose PsyRoom B for generating personalized treatment plans. Quantitative evaluations demonstrate that AI PsyRoom significantly outperforms state-of-the-art methods, achieving 18% improvement in problem orientation, 23% in expression, 24% in Empathy, and 16% in interactive communication quality. The datasets and models are publicly available, providing a foundation for advancing AI-assisted psychological counseling research.

cs.AI↗

arXiv2025

SmartThinker: Learning to Compress and Preserve Reasoning by Step-Level Length Control

Large reasoning models (LRMs) have exhibited remarkable reasoning capabilities through inference-time scaling, but this progress has also introduced considerable redundancy and inefficiency into their reasoning processes, resulting in substantial computational waste. Previous work has attempted to mitigate this issue by penalizing the overall length of generated samples during reinforcement learning (RL), with the goal of encouraging a more concise chains of thought. However, we observe that such global length penalty often lead to excessive compression of critical reasoning steps while preserving unnecessary details in simpler ones, yielding a suboptimal trade-off between accuracy and efficiency. To address this issue, we propose SmartThinker, a two-stage learnable framework designed to enable fine-grained control over the length of reasoning chains based on the importance of each individual step. In the first stage, SmartThinker adapts a reasoning model to a short-form reasoning mode through rejection sampling combined with supervised fine-tuning (SFT). In the second stage, SmartThinker applies Step-Level Length Control Policy Optimization (SCPO) to refine the model output distribution, which increases the proportion of length allocated to critical steps while reducing redundancy in less important ones. SCPO consists of four core components: an online importance estimator, a step-level length control reward function, a step-level generalized advantage estimation (S-GAE) and a difficulty-adaptive clipping strategy. Working in concert, these components enable SCPO to implement differentiated length control across reasoning steps. Empirical results across multiple reasoning benchmarks and various backbone models demonstrate that SmartThinker significantly reduces redundant reasoning while achieving comparable or even superior performance to existing methods.

cs.AI↗

arXiv2025

Test-time Adaptation for Foundation Medical Segmentation Model without Parametric Updates

Foundation medical segmentation models, with MedSAM being the most popular, have achieved promising performance across organs and lesions. However, MedSAM still suffers from compromised performance on specific lesions with intricate structures and appearance, as well as bounding box prompt-induced perturbations. Although current test-time adaptation (TTA) methods for medical image segmentation may tackle this issue, partial (e.g., batch normalization) or whole parametric updates restrict their effectiveness due to limited update signals or catastrophic forgetting in large models. Meanwhile, these approaches ignore the computational complexity during adaptation, which is particularly significant for modern foundation models. To this end, our theoretical analyses reveal that directly refining image embeddings is feasible to approach the same goal as parametric updates under the MedSAM architecture, which enables us to realize high computational efficiency and segmentation performance without the risk of catastrophic forgetting. Under this framework, we propose to encourage maximizing factorized conditional probabilities of the posterior prediction probability using a proposed distribution-approximated latent conditional random field loss combined with an entropy minimization loss. Experiments show that we achieve about 3\% Dice score improvements across three datasets while reducing computational complexity by over 7 times.

q-bio.QM↗

arXiv2025

Preserving spreading dynamics and information flow in complex network reduction

Effectively preserving both the structural and dynamical properties during the reduction of complex networks remains a significant research topic. Existing network reduction methods based on renormalization group or sampling often face challenges such as high computational complexity and the loss of critical dynamic attributes. This paper proposes an efficient network reduction framework based on subgraph extraction, which accurately preserves epidemic spreading dynamics and information flow through a coordinated optimization strategy of node removal and edge pruning. Specifically, a node removal algorithm driven by enhanced degree centrality is introduced to preferentially remove low-centrality nodes, thereby constructing a smaller-scale subnetwork. Subsequently, an edge pruning algorithm is designed to regulate the edge density of the subnetwork, ensuring that its average degree remains approximately consistent with that of the original network. Experimental results on Erdös-Rényi random graphs, Barabási-Albert scale-free networks, and real-world social contact networks from various domains demonstrate that this proposed method can reduce the size of networks with heterogeneous structures by more than 85\%, while preserving their epidemic dynamics and information flow. More importantly, our method almost always achieves the highest accuracy compared to state-of-the-art techniques. These findings provide valuable insights for predicting the dynamical behavior of large-scale real-world networks, and also reveal that a large number of nodes and edges in real-world networks play redundant roles in information transmission.

cs.SI↗

arXiv2025

Goal-Oriented Sequential Bayesian Experimental Design for Causal Learning

We present GO-CBED, a goal-oriented Bayesian framework for sequential causal experimental design. Unlike conventional approaches that select interventions aimed at inferring the full causal model, GO-CBED directly maximizes the expected information gain (EIG) on user-specified causal quantities of interest, enabling more targeted and efficient experimentation. The framework is both non-myopic, optimizing over entire intervention sequences, and goal-oriented, targeting only model aspects relevant to the causal query. To address the intractability of exact EIG computation, we introduce a variational lower bound estimator, optimized jointly through a transformer-based policy network and normalizing flow-based variational posteriors. The resulting policy enables real-time decision-making via an amortized network. We demonstrate that GO-CBED consistently outperforms existing baselines across various causal reasoning and discovery tasks-including synthetic structural causal models and semi-synthetic gene regulatory networks-particularly in settings with limited experimental budgets and complex causal mechanisms. Our results highlight the benefits of aligning experimental design objectives with specific research goals and of forward-looking sequential planning.

cs.LG↗

arXiv2025

AbdomenAtlas-8K: Annotating 8,000 CT Volumes for Multi-Organ Segmentation in Three Weeks

Annotating medical images, particularly for organ segmentation, is laborious and time-consuming. For example, annotating an abdominal organ requires an estimated rate of 30-60 minutes per CT volume based on the expertise of an annotator and the size, visibility, and complexity of the organ. Therefore, publicly available datasets for multi-organ segmentation are often limited in data size and organ diversity. This paper proposes an active learning method to expedite the annotation process for organ segmentation and creates the largest multi-organ dataset (by far) with the spleen, liver, kidneys, stomach, gallbladder, pancreas, aorta, and IVC annotated in 8,448 CT volumes, equating to 3.2 million slices. The conventional annotation methods would take an experienced annotator up to 1,600 weeks (or roughly 30.8 years) to complete this task. In contrast, our annotation method has accomplished this task in three weeks (based on an 8-hour workday, five days a week) while maintaining a similar or even better annotation quality. This achievement is attributed to three unique properties of our method: (1) label bias reduction using multiple pre-trained segmentation models, (2) effective error detection in the model predictions, and (3) attention guidance for annotators to make corrections on the most salient errors. Furthermore, we summarize the taxonomy of common errors made by AI algorithms and annotators. This allows for continuous revision of both AI and annotations and significantly reduces the annotation costs required to create large-scale datasets for a wider variety of medical imaging tasks.

eess.IV↗

arXiv2025

Simulation studies on compensation for incoherent magnet error driven half-integer and 3rd-order resonances with space charge in HIAF-BRing

Magnet error driven resonances under space-charge-induced resonance crossing represents a significant limitation on beam intensity in low-injection-energy high-intensity synchrotrons. Unstable particle motion arises from the combined effects of space-charge-induced tune spread and magnetic field imperfections, which limits the intensity. This paper introduces a space-charge-Twiss modification to the Resonance Driving Terms (RDTs). The new RDTs are named modified RDTs. Modified RDTs aim to describe the nonlinear behavior under incoherent error driven resonances induced by the combined effect of space charge and magnetic field imperfections. The feasibility of the compensation scheme with modified RDTs is demonstrated through coasting-beam simulations under space-charge-induced half-integer and 3rd-order resonances, using the lattice of the High Intensity Heavy-Ion Accelerator Facility Booster Ring (HIAF-BRing). The simulations demonstrate that the compensation through minimizing modified RDTs significantly suppresses the emittance growth and the unstable particle motion. Besides, the effect of compensation scheme with modified RDTs against periodical resonance crossing is evaluated in bunched-beam simulations.

physics.acc-ph↗

arXiv2025

Numerical simulations of electron acceleration driven by heavy ion beams in plasma with alternating density gradients

Plasma-Based Acceleration (PBA) has been demonstrated using laser, electron, and proton drivers. However, significant challenges remain in achieving high efficiency, stable acceleration, and scalable energy gain. Heavy ion beam drivers, with their high kinetic energy, offer the potential for greater energy transfer to the witness beam. Unfortunately, limited by the relatively low velocity of heavy ion, the dephasing length is really short leading to a low energy gain of the witness beam. Conventional method that increase plasma density linearly is ineffective in this context because the mismatch between the RMS beam radius and plasma wavelength will make the wakefield degrade or even disappear. In this paper, we propose a method that periodically switches the witness beam between different accelerating phase, allowing it to shift between adjacent accelerating cavities. Therefore, the plasma density does not only strictly increase, but also decrease. This will help maintain the structure of wakefield and increase the energy gain of the witness beam.

physics.acc-ph↗

arXiv2025

Probabilistic Prototype Calibration of Vision-Language Models for Generalized Few-shot Semantic Segmentation

Generalized Few-Shot Semantic Segmentation (GFSS) aims to extend a segmentation model to novel classes with only a few annotated examples while maintaining performance on base classes. Recently, pretrained vision-language models (VLMs) such as CLIP have been leveraged in GFSS to improve generalization on novel classes through multi-modal prototypes learning. However, existing prototype-based methods are inherently deterministic, limiting the adaptability of learned prototypes to diverse samples, particularly for novel classes with scarce annotations. To address this, we propose FewCLIP, a probabilistic prototype calibration framework over multi-modal prototypes from the pretrained CLIP, thus providing more adaptive prototype learning for GFSS. Specifically, FewCLIP first introduces a prototype calibration mechanism, which refines frozen textual prototypes with learnable visual calibration prototypes, leading to a more discriminative and adaptive representation. Furthermore, unlike deterministic prototype learning techniques, FewCLIP introduces distribution regularization over these calibration prototypes. This probabilistic formulation ensures structured and uncertainty-aware prototype learning, effectively mitigating overfitting to limited novel class data while enhancing generalization. Extensive experimental results on PASCAL-5$^i$ and COCO-20$^i$ datasets demonstrate that our proposed FewCLIP significantly outperforms state-of-the-art approaches across both GFSS and class-incremental setting. The code is available at https://github.com/jliu4ai/FewCLIP.

cs.CV↗

arXiv2025

A corresponding relationship between nonlinear Hermitian systems and linear non-Hermitian models

We note that the non-orthogonality of states and their coincidence at the degeneracy point are both admitted by nonlinear Hermitian systems and linear non-Hermitian systems. These striking characteristics motivate us to re-investigate the localized waves of nonlinear Hermitian systems and the eigenvalue degeneracies of linear non-Hermitian models, based on several well-known Lax integrable systems that have wide applications in nonlinear optics. We choose nonlinear Schrodinger equation integrability hierarchy to demonstrate the quantitative relations between dynamics of nonlinear Hermitian systems and eigenvalue degeneracies of linear non-Hermitian models. Specifically, the degeneracies of the real or imaginary spectrum of the linear non-Hermitian matrices are uncovered to clarify several essential characteristics of nonlinear localized waves, such as breathers, rogue waves, and solitons. We find that the exceptional points generally correspond to rogue waves for modulational instability cases and dark solitons with maximum velocity for the modulational stability cases. These insights provide another interesting perspective for understanding nonlinear localized waves, and hint that there are closer relations between nonlinear Hermitian systems and linear non-Hermitian systems.

nlin.PS↗