arXiv subjects

Jie Liu

Publications and source records attributed to Jie Liu.

At least 217 records · Page 12Linked to original sources

arXiv2025

CDT: A Comprehensive Capability Framework for Large Language Models Across Cognition, Domain, and Task

Recent advances in Large Language Models (LLMs) have significantly enhanced their capabilities, highlighting the need for comprehensive evaluation frameworks that extend beyond task-specific benchmarks. However, existing benchmarks often focus on isolated abilities, lacking a holistic framework for assessing LLM capabilities. To address this gap, we propose the Cognition-Domain-Task (CDT) framework, which comprehensively measures a model's capabilities across three dimensions. We expand the scope of model capability definitions at the cognitive level by incorporating the Cattell-Horn-Carroll cognitive theory, refining the categorization of model capabilities. We apply CDT in two directions: dataset capability evaluation and data selection. Experiments show that our capability metrics correlate well with downstream performance and can support effective dataset analysis and construction. The experiments on data selection also show significant improvements in both general and specific benchmarks, achieving scores of 44.3 and 45.4, with an increase of 1.6 and 2.2 points over the baselines, respectively. These results validate the effectiveness and practicality of CDT. Source code and models are available at https://github.com/Alessa-mo/CDT.

cs.CL↗

arXiv2025

Rich unconventional Hall effects in a single quasi-kagome Kondo Weyl semimetal candidate Ce$_3$TiSb$_5$

It is generally believed that electronic correlation, geometric frustration, and topology, \textit{individually}, can facilitate the emergence of various intriguing properties that have attracted a broad audience for both fundamental research and potential applications. Here, we report a series of unconventional Hall effects observed in a \textit{single} compound - quasi-kagome Kondo Weyl semimetal candidate Ce$_3$TiSb$_5$. In the paramagnetic phase, signature of dynamic $c$-$f$ hybridization is revealed by a reduction of anomalous Hall effect and is connected to frustration-promoted incoherent Kondo scattering. A large topological Hall effect exceeding 0.2 $μΩ\cdot$cm is found at low temperatures, which should be ascribed to the non-collinear magnetic texture. In addition, a peculiar loop-shaped Hall effect with switching chirality is also seen, which is inferred to be associated with magnetic domain walls that pin history-dependent spin chirality and / or Fermi-arc surface states projected from the in-gap Weyl nodes. These exotic results place Ce$_3$TiSb$_5$ in a regime of highly-frustrated antiferromagnetic dense Kondo lattice with a nontrivial topology on an ``extended" global phase diagram, and highlight the interplay among electronic correlation, geometric frustration and topology.

cond-mat.str-el↗

arXiv2025

DS-Diffusion: Data Style-Guided Diffusion Model for Time-Series Generation

Diffusion models are the mainstream approach for time series generation tasks. However, existing diffusion models for time series generation require retraining the entire framework to introduce specific conditional guidance. There also exists a certain degree of distributional bias between the generated data and the real data, which leads to potential model biases in downstream tasks. Additionally, the complexity of diffusion models and the latent spaces leads to an uninterpretable inference process. To address these issues, we propose the data style-guided diffusion model (DS-Diffusion). In the DS-Diffusion, a diffusion framework based on style-guided kernels is developed to avoid retraining for specific conditions. The time-information based hierarchical denoising mechanism (THD) is developed to reduce the distributional bias between the generated data and the real data. Furthermore, the generated samples can clearly indicate the data style from which they originate. We conduct comprehensive evaluations using multiple public datasets to validate our approach. Experimental results show that, compared to the state-of-the-art model such as ImagenTime, the predictive score and the discriminative score decrease by 5.56% and 61.55%, respectively. The distributional bias between the generated data and the real data is further reduced, the inference process is also more interpretable. Moreover, by eliminating the need to retrain the diffusion model, the flexibility and adaptability of the model to specific conditions are also enhanced.

cs.LG↗

arXiv2025

Quantum phase transition in two dimension nonlinear cavity optomagnonic system

The superfluid-Mott insulator and ergodic-many body localization transitions based on a two dimension nonlinear cavity optomagnonic system are investigated, where a yttrium iron garnet (YIG) sphere is embedded at each site of a two-dimensional coupled cavity array. It can be demonstrated that, the introduction of phonon-photon coupling enhances the coherence of the system when considering the Kerr nonlinearity of the YIG sphere. In contrast, the Kerr nonlinearity of photons is more conducive to the Mott insulating phase than that of the YIG sphere. We further elucidate the underlying physical mechanism by calculating the effective repulsive potential of the system in the presence of photonic Kerr nonlinearity. Regarding the ergodic-many body localization transition, the results indicate that as the disorder strength of the Kerr nonlinearity increases, the system transitions from the ergodic phase to the many body localized phase, while increasing the chemical potential expands the region of the ergodic phase. This work provides a novel framework for characterizing quantum phase transitions in cavity optomagnonic systems and offers an experimentally feasible scheme for studying them, thereby yielding valuable insights for quantum simulation.

quant-ph↗

arXiv2025

Symplectic singularities arising from algebras of symmetric tensors

The algebra of symmetric tensors $S(X):= H^0(X, \sf{S}^{\bullet} T_X)$ of a projective manifold $X$ leads to a natural dominant affinization morphism $$ φ_X: T^*X \longrightarrow \mathcal{Z}_X:= \text{Spec} S(X). $$ It is shown that $φ_X$ is birational if and only if $T_X$ is big. We prove that if $φ_X$ is birational, then $\mathcal{Z}_X$ is a symplectic variety endowed with the Schouten--Nijenhuis bracket if and only if $\mathbb{P} T_X$ is of Fano type, which is the case for smooth projective toric varieties, smooth horospherical varieties with small boundary and the quintic del Pezzo threefold. These give examples of a distinguished class of conical symplectic varieties, which we call symplectic orbifold cones.

math.AG↗

arXiv2025

Anomalous Trajectory Drift and Geometric Phases of Cyclic Spinor Solitons Induced by Virtual Magnetic Monopoles

We investigate the dynamics of a two-component Bose-Einstein condensate with spin-orbit coupling numerically and analytically. Under the drive of a weak segmented rotational external field, we observe that the system exhibits cyclic soliton motion; however, in contrast to the predictions of quasi-particle theory, the trajectory of the soliton center shows a distinct drift. The underlying mechanism of this anomalous drift is revealed: the moving soliton experiences a Lorentz force induced by a virtual magnetic monopole field in momentum space. We further calculate the phase evolution of the soliton during this cyclic motion and find that its geometric component comprises both an adiabatic Berry phase and a nonadiabatic Aharonov-Anandan phase. Notably, the Berry phase can be expressed in terms of the magnetic flux of the aforementioned virtual monopole field. Our findings hold implications for geometric phase theory and experiments on two-component Bose-Einstein condensates, and may establish a novel link between quantum geometry and soliton dynamics.

cond-mat.quant-gas↗

arXiv2025

Hunyuan3D Studio: End-to-End AI Pipeline for Game-Ready 3D Asset Generation

The creation of high-quality 3D assets, a cornerstone of modern game development, has long been characterized by labor-intensive and specialized workflows. This paper presents Hunyuan3D Studio, an end-to-end AI-powered content creation platform designed to revolutionize the game production pipeline by automating and streamlining the generation of game-ready 3D assets. At its core, Hunyuan3D Studio integrates a suite of advanced neural modules (such as Part-level 3D Generation, Polygon Generation, Semantic UV, etc.) into a cohesive and user-friendly system. This unified framework allows for the rapid transformation of a single concept image or textual description into a fully-realized, production-quality 3D model complete with optimized geometry and high-fidelity PBR textures. We demonstrate that assets generated by Hunyuan3D Studio are not only visually compelling but also adhere to the stringent technical requirements of contemporary game engines, significantly reducing iteration time and lowering the barrier to entry for 3D content creation. By providing a seamless bridge from creative intent to technical asset, Hunyuan3D Studio represents a significant leap forward for AI-assisted workflows in game development and interactive media.

cs.CV↗

arXiv2025

Special Issue: Commemorating the 110th Anniversary of TANG Au-chin's Birthday Calculation of the Green's function on near-term quantum computers via Cartan decomposition

Accurate computation of the Green's function is crucial for connecting experimental observations to the underlying quantum states. A major challenge in evaluating the Green's function in the time domain lies in the efficient simulation of quantum state evolution under a given Hamiltonian-a task that becomes exponentially complex for strongly correlated systems on classical computers. Quantum computing provides a promising pathway to overcome this barrier by enabling efficient simulation of quantum dynamics. However, for near-term quantum devices with limited coherence times and fidelity, the deep quantum circuits required to implement time-evolution operators present a significant challenge for practical applications. In this work, we introduce an efficient algorithm for computing Green's functions via Cartan decomposition, which requires only fixed-depth quantum circuits for arbitrarily long time simulations. Additionally, analytical gradients are formulated to accelerate the Cartan decomposition by leveraging a unitary transformation in a factorized form. The new algorithm is applied to simulate long-time Green's functions for the Fermi-Hubbard and transverse-field Ising models, extracting the spectral functions through Fourier transformation.

quant-ph↗

arXiv2025

RewardDance: Reward Scaling in Visual Generation

Reward Models (RMs) are critical for improving generation models via Reinforcement Learning (RL), yet the RM scaling paradigm in visual generation remains largely unexplored. It primarily due to fundamental limitations in existing approaches: CLIP-based RMs suffer from architectural and input modality constraints, while prevalent Bradley-Terry losses are fundamentally misaligned with the next-token prediction mechanism of Vision-Language Models (VLMs), hindering effective scaling. More critically, the RLHF optimization process is plagued by Reward Hacking issue, where models exploit flaws in the reward signal without improving true quality. To address these challenges, we introduce RewardDance, a scalable reward modeling framework that overcomes these barriers through a novel generative reward paradigm. By reformulating the reward score as the model's probability of predicting a "yes" token, indicating that the generated image outperforms a reference image according to specific criteria, RewardDance intrinsically aligns reward objectives with VLM architectures. This alignment unlocks scaling across two dimensions: (1) Model Scaling: Systematic scaling of RMs up to 26 billion parameters; (2) Context Scaling: Integration of task-specific instructions, reference examples, and chain-of-thought (CoT) reasoning. Extensive experiments demonstrate that RewardDance significantly surpasses state-of-the-art methods in text-to-image, text-to-video, and image-to-video generation. Crucially, we resolve the persistent challenge of "reward hacking": Our large-scale RMs exhibit and maintain high reward variance during RL fine-tuning, proving their resistance to hacking and ability to produce diverse, high-quality outputs. It greatly relieves the mode collapse problem that plagues smaller models.

cs.CV↗

arXiv2025

KLIPA: A Knowledge Graph and LLM-Driven QA Framework for IP Analysis

Effectively managing intellectual property is a significant challenge. Traditional methods for patent analysis depend on labor-intensive manual searches and rigid keyword matching. These approaches are often inefficient and struggle to reveal the complex relationships hidden within large patent datasets, hindering strategic decision-making. To overcome these limitations, we introduce KLIPA, a novel framework that leverages a knowledge graph and a large language model (LLM) to significantly advance patent analysis. Our approach integrates three key components: a structured knowledge graph to map explicit relationships between patents, a retrieval-augmented generation(RAG) system to uncover contextual connections, and an intelligent agent that dynamically determines the optimal strategy for resolving user queries. We validated KLIPA on a comprehensive, real-world patent database, where it demonstrated substantial improvements in knowledge extraction, discovery of novel connections, and overall operational efficiency. This combination of technologies enhances retrieval accuracy, reduces reliance on domain experts, and provides a scalable, automated solution for any organization managing intellectual property, including technology corporations and legal firms, allowing them to better navigate the complexities of strategic innovation and competitive intelligence.

cs.IR↗

arXiv2025

AIM 2025 Challenge on High FPS Motion Deblurring: Methods and Results

This paper presents a comprehensive review of the AIM 2025 High FPS Non-Uniform Motion Deblurring Challenge, highlighting the proposed solutions and final results. The objective of this challenge is to identify effective networks capable of producing clearer and visually compelling images in diverse and challenging conditions, by learning representative visual cues for complex aggregations of motion types. A total of 68 participants registered for the competition, and 9 teams ultimately submitted valid entries. This paper thoroughly evaluates the state-of-the-art advances in high-FPS single image motion deblurring, showcasing the significant progress in the field, while leveraging samples of the novel dataset, MIORe, that introduces challenging examples of movement patterns.

cs.CV↗

arXiv2025

Decoupled Entity Representation Learning for Pinterest Ads Ranking

In this paper, we introduce a novel framework following an upstream-downstream paradigm to construct user and item (Pin) embeddings from diverse data sources, which are essential for Pinterest to deliver personalized Pins and ads effectively. Our upstream models are trained on extensive data sources featuring varied signals, utilizing complex architectures to capture intricate relationships between users and Pins on Pinterest. To ensure scalability of the upstream models, entity embeddings are learned, and regularly refreshed, rather than real-time computation, allowing for asynchronous interaction between the upstream and downstream models. These embeddings are then integrated as input features in numerous downstream tasks, including ad retrieval and ranking models for CTR and CVR predictions. We demonstrate that our framework achieves notable performance improvements in both offline and online settings across various downstream tasks. This framework has been deployed in Pinterest's production ad ranking systems, resulting in significant gains in online metrics.

cs.IR↗

arXiv2025

Inference-Time Alignment Control for Diffusion Models with Reinforcement Learning Guidance

Denoising-based generative models, particularly diffusion and flow matching algorithms, have achieved remarkable success. However, aligning their output distributions with complex downstream objectives, such as human preferences, compositional accuracy, or data compressibility, remains challenging. While reinforcement learning (RL) fine-tuning methods, inspired by advances in RL from human feedback (RLHF) for large language models, have been adapted to these generative frameworks, current RL approaches are suboptimal for diffusion models and offer limited flexibility in controlling alignment strength after fine-tuning. In this work, we reinterpret RL fine-tuning for diffusion models through the lens of stochastic differential equations and implicit reward conditioning. We introduce Reinforcement Learning Guidance (RLG), an inference-time method that adapts Classifier-Free Guidance (CFG) by combining the outputs of the base and RL fine-tuned models via a geometric average. Our theoretical analysis shows that RLG's guidance scale is mathematically equivalent to adjusting the KL-regularization coefficient in standard RL objectives, enabling dynamic control over the alignment-quality trade-off without further training. Extensive experiments demonstrate that RLG consistently improves the performance of RL fine-tuned models across various architectures, RL algorithms, and downstream tasks, including human preferences, compositional control, compressibility, and text rendering. Furthermore, RLG supports both interpolation and extrapolation, thereby offering unprecedented flexibility in controlling generative alignment. Our approach provides a practical and theoretically sound solution for enhancing and controlling diffusion model alignment at inference. The source code for RLG is publicly available at the Github: https://github.com/jinluo12345/Reinforcement-learning-guidance.

cs.LG↗

arXiv2025

Long-term simultaneous 2.25/8.60~GHz monitoring of the newly-discovered repeating FRB~20240114A

We report on the simultaneous monitoring of the repeating fast radio burst (FRB) 20240114A at 2.25 and 8.60~GHz, conducted 66 times between 2024 January 29 and 2025 February 15 with the Shanghai Tianma Radio Telescope (TMRT). In about 180 hours of observation, we detected 155 bursts at 2.25~GHz above a fluence threshold of 0.72~Jy~ms, but none at 8.60~GHz above a fluence threshold of 0.27~Jy~ms. FRB~20240114A exhibited frequency-dependent activity, as evidenced by the non-detections in 14.3 hours of observations at 2.25~GHz prior to 2024 February 24, despite its reported activity below 2~GHz. In contrast to its low-activity state reported below 1.4~GHz between 2024 June and December, FRB~20240114A exhibited high activity at 2.25~GHz in 2024 July with a mean burst rate of $1.72^{+0.18}_{-0.16}~\rm{hr}^{-1}$, followed by a low-activity state. We also detected a short-term reactivation at 2.25~GHz around 2025 January 20, about two weeks after renewed activity was reported below 1.4~GHz by other telescopes. The median burst width at 2.25~GHz is 3~ms, which is narrower than that at lower frequencies. The waiting time distribution peaks at 1019~s, and burst arrivals on hourly timescales consistent with a Poisson process. The isotropic-equivalent energy of bursts spans $10^{37} -10^{39}$~erg. The distribution of burst energy above the completeness threshold ($7.5\times10^{37}$~erg) follows a power-law relation with an index of $γ=-1.20\pm0.03\pm0.02$. Finally, we find that FRB~20240114A is at least two orders of magnitude less active at 8.60~GHz than at 2.25~GHz, and we constrain the broadband spectra of the detected bursts.

astro-ph.HE↗

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↗