arXiv subjects

Kai Liu

Publications and source records attributed to Kai Liu.

At least 109 records · Page 6Linked to original sources

arXiv2026

La RoSA: Enhancing LLM Efficiency via Layerwise Rotated Sparse Activation

Activation sparsity can reduce the computational overhead and memory transfers during the forward pass of Large Language Model (LLM) inference. Existing methods face limitations, either demanding time-consuming recovery training that hinders real-world adoption, or relying on empirical magnitude-based pruning, which causes fluctuating sparsity and unstable inference speed-up. This paper introduces LaRoSA (Layerwise Rotated Sparse Activation), a novel method for activation sparsification designed to improve LLM efficiency without requiring additional training or magnitude-based pruning. We leverage layerwise orthogonal rotations to transform input activations into rotated forms that are more suitable for sparsification. By employing a Top-K selection approach within the rotated activations, we achieve consistent model-level sparsity and reliable wall-clock time speed-up. LaRoSA is effective across various sizes and types of LLMs, demonstrating minimal performance degradation and robust inference acceleration. Specifically, for LLaMA2-7B at 40% sparsity, LaRoSA achieves a mere 0.17 perplexity gap with a consistent 1.30x wall-clock time speed-up, and reduces the accuracy gap in zero-shot tasks compared to the dense model to just 0.54%, while surpassing TEAL by 1.77% and CATS by 17.14%.

cs.CL↗

arXiv2026

JavisGPT: A Unified Multi-modal LLM for Sounding-Video Comprehension and Generation

This paper presents JavisGPT, the first unified multimodal large language model (MLLM) for joint audio-video (JAV) comprehension and generation. JavisGPT has a concise encoder-LLM-decoder architecture, which has a SyncFusion module for spatio-temporal audio-video fusion and synchrony-aware learnable queries to bridge a pretrained JAV-DiT generator. This design enables temporally coherent video-audio understanding and generation from multimodal instructions. We design an effective three-stage training pipeline consisting of multimodal pretraining, audio-video fine-tuning, and large-scale instruction-tuning, to progressively build multimodal comprehension and generation from existing vision-language models. For instruction tuning, we construct JavisInst-Omni, a high-quality instruction dataset with over 200K GPT-4o-curated audio-video-text dialogues that cover diverse and multi-level comprehension and generation scenarios. On JAV comprehension and generation benchmarks, our experiments show that JavisGPT outperforms existing MLLMs, particularly in complex and temporally synchronized settings.

cs.CV↗

arXiv2025

Nearly-room-temperature ferromagnetism and tunable anomalous Hall effect in atomically thin Fe4CoGeTe2

Itinerant ferromagnetism at room temperature is a key ingredient for spin transport and manipulation. Here, we report the realization of nearly-room-temperature itinerant ferromagnetism in Co doped Fe5GeTe2 thin flakes. The ferromagnetic transition temperature TC (~ 323 K - 337 K) is almost unchanged when thickness is down to 12 nm and is still about 284 K at 2 nm (bilayer thickness). Theoretical calculations further indicate that the ferromagnetism persists in monolayer Fe4CoGeTe2. In addition to the robust ferromagnetism down to the ultrathin limit, Fe4CoGeTe2 exhibits an unusual temperature- and thickness-dependent intrinsic anomalous Hall effect. We propose that it could be ascribed to the dependence of band structure on thickness that changes the Berry curvature near the Fermi energy level subtly. The nearly-room-temperature ferromagnetism and tunable anomalous Hall effect in atomically thin Fe4CoGeTe2 provide opportunities to understand the exotic transport properties of two-dimensional van der Waals magnetic materials and explore their potential applications in spintronics.

cond-mat.mtrl-sci↗

arXiv2025

UmniBench: Unified Understand and Generation Model Oriented Omni-dimensional Benchmark

Unifying multimodal understanding and generation has shown impressive capabilities in cutting-edge proprietary systems. However, evaluations of unified multimodal models (UMMs) remain decoupled, assessing their understanding and generation abilities separately with corresponding datasets. To address this, we propose UmniBench, a benchmark tailored for UMMs with omni-dimensional evaluation. First, UmniBench can assess the understanding, generation, and editing ability within a single evaluation process. Based on human-examined prompts and QA pairs, UmniBench leverages UMM itself to evaluate its generation and editing ability with its understanding ability. This simple but effective paradigm allows comprehensive evaluation of UMMs. Second, UmniBench covers 13 major domains and more than 200 concepts, ensuring a thorough inspection of UMMs. Moreover, UmniBench can also decouple and separately evaluate understanding, generation, and editing abilities, providing a fine-grained assessment. Based on UmniBench, we benchmark 24 popular models, including both UMMs and single-ability large models. We hope this benchmark provides a more comprehensive and objective view of unified models and logistical support for improving the performance of the community model.

cs.AI↗

arXiv2025

Fose: Fusion of One-Step Diffusion and End-to-End Network for Pansharpening

Pansharpening is a significant image fusion task that fuses low-resolution multispectral images (LRMSI) and high-resolution panchromatic images (PAN) to obtain high-resolution multispectral images (HRMSI). The development of the diffusion models (DM) and the end-to-end models (E2E model) has greatly improved the frontier of pansharping. DM takes the multi-step diffusion to obtain an accurate estimation of the residual between LRMSI and HRMSI. However, the multi-step process takes large computational power and is time-consuming. As for E2E models, their performance is still limited by the lack of prior and simple structure. In this paper, we propose a novel four-stage training strategy to obtain a lightweight network Fose, which fuses one-step DM and an E2E model. We perform one-step distillation on an enhanced SOTA DM for pansharping to compress the inference process from 50 steps to only 1 step. Then we fuse the E2E model with one-step DM with lightweight ensemble blocks. Comprehensive experiments are conducted to demonstrate the significant improvement of the proposed Fose on three commonly used benchmarks. Moreover, we achieve a 7.42 speedup ratio compared to the baseline DM while achieving much better performance. The code and model are released at https://github.com/Kai-Liu001/Fose.

cs.CV↗

arXiv2025

RoFt-Mol: Benchmarking Robust Fine-Tuning with Molecular Graph Foundation Models

In the era of foundation models, fine-tuning pre-trained models for specific downstream tasks has become crucial. This drives the need for robust fine-tuning methods to address challenges such as model overfitting and sparse labeling. Molecular graph foundation models (MGFMs) face unique difficulties that complicate fine-tuning. These models are limited by smaller pre-training datasets and more severe data scarcity for downstream tasks, both of which require enhanced model generalization. Moreover, MGFMs must accommodate diverse objectives, including both regression and classification tasks. To better understand and improve fine-tuning techniques under these conditions, we classify eight fine-tuning methods into three mechanisms: weight-based, representation-based, and partial fine-tuning. We benchmark these methods on downstream regression and classification tasks across supervised and self-supervised pre-trained models in diverse labeling settings. This extensive evaluation provides valuable insights and informs the design of a refined robust fine-tuning method, ROFT-MOL. This approach combines the strengths of simple post-hoc weight interpolation with more complex weight ensemble fine-tuning methods, delivering improved performance across both task types while maintaining the ease of use inherent in post-hoc weight interpolation.

cs.LG↗

arXiv2025

HunyuanOCR Technical Report

This paper presents HunyuanOCR, a commercial-grade, open-source, and lightweight (1B parameters) Vision-Language Model (VLM) dedicated to OCR tasks. The architecture comprises a Native Vision Transformer (ViT) and a lightweight LLM connected via an MLP adapter. HunyuanOCR demonstrates superior performance, outperforming commercial APIs, traditional pipelines, and larger models (e.g., Qwen3-VL-4B). Specifically, it surpasses current public solutions in perception tasks (Text Spotting, Parsing) and excels in semantic tasks (IE, Text Image Translation), securing first place in the ICDAR 2025 DIMT Challenge (Small Model Track). Furthermore, it achieves state-of-the-art (SOTA) results on OCRBench among VLMs with fewer than 3B parameters. HunyuanOCR achieves breakthroughs in three key aspects: 1) Unifying Versatility and Efficiency: We implement comprehensive support for core capabilities including spotting, parsing, IE, VQA, and translation within a lightweight framework. This addresses the limitations of narrow "OCR expert models" and inefficient "General VLMs". 2) Streamlined End-to-End Architecture: Adopting a pure end-to-end paradigm eliminates dependencies on pre-processing modules (e.g., layout analysis). This fundamentally resolves error propagation common in traditional pipelines and simplifies system deployment. 3) Data-Driven and RL Strategies: We confirm the critical role of high-quality data and, for the first time in the industry, demonstrate that Reinforcement Learning (RL) strategies yield significant performance gains in OCR tasks. HunyuanOCR is officially open-sourced on HuggingFace. We also provide a high-performance deployment solution based on vLLM, placing its production efficiency in the top tier. We hope this model will advance frontier research and provide a solid foundation for industrial applications.

cs.CV↗

arXiv2025

Robust Phase-Shifting Profilometry for Arbitrary Motion

Phase-shifting profilometry (PSP) enables high-accuracy 3D reconstruction but remains highly susceptible to object motion. Although numerous studies have explored compensation for motion-induced errors, residual inaccuracies still persist, particularly in complex motion scenarios. In this paper, we propose a robust phase-shifting profilometry for arbitrary motion (RPSP-AM), including six-degrees-of-freedom (6-DoF) motion (translation and rotation in any direction), non-rigid deformations, and multi-target movements, achieving high-fidelity motion-error-free 3D reconstruction. We categorize motion errors into two components: 1) ghosting artifacts induced by image misalignment, and 2) ripple-like distortions induced by phase deviation. To eliminate the ghosting artifacts, we perform pixel-wise image alignment based on dense optical flow tracking. To correct ripple-like distortions, we propose a high-accuracy, low-complexity image-sequential binomial self-compensation (I-BSC) method, which performs a summation of the homogeneous fringe images weighted by binomial coefficients, exponentially reducing the ripple-like distortions with a competitive computational speed compared with the traditional four-step phase-shifting method. Extensive experimental results demonstrate that, under challenging conditions such as 6-DoF motion, non-rigid deformations, and multi-target movements, the proposed RPSP-AM outperforms state-of-the-art (SoTA) methods in compensating for both ghosting artifacts and ripple-like distortions. Our approach extends the applicability of PSP to arbitrary motion scenarios, endowing it with potential for widespread adoption in fields such as robotics, industrial inspection, and medical reconstruction.

cs.CV↗

arXiv2025

Coexistence of near-EF van Hove singularity and in-gap topological Dirac surface states in superconducting electrides

Superconducting electrides have attracted growing attention for their potential to achieve high superconducting transition temperatures (TC) under pressure. However, many known electrides are chemically reactive and unstable, making high-quality single-crystal growth, characterization, and measurements difficult, and most do not exhibit superconductivity at ambient pressure. In contrast, La3In stands out for its ambient-pressure superconductivity (TC ~ 9.4 K) and the availability of high-quality single crystals. Here, we investigate its low-energy electronic structure using angle-resolved photoemission spectroscopy and first-principles calculations. The bands near the Fermi energy are mainly derived from La 5d and In 5p orbitals. A saddle point is directly observed at the Brillouin zone (BZ) boundary, while a three-dimensional van Hove singularity crosses EF at the BZ corner. First-principles calculations further reveal topological Dirac surface states within the bulk energy gap above EF. The coexistence of a high density of states and in-gap topological surface states near EF suggests that La3In offers a promising platform for tuning superconductivity and exploring possible topological superconducting phases through doping or external pressure.

cond-mat.supr-con↗

arXiv2025

Pinching-Antenna Systems-Assisted SWIPT: A Rate-Energy Trade-off Perspective

This paper investigates the rate-energy trade-off for pinching-antenna systems (PASS)-assisted simultaneous wireless information and power transfer (SWIPT) systems. Both the single information user (IU)/energy user (EU) and multiple IUs/EUs scenarios are considered.1) For the single IU/EU scenario, a pinching beamforming optimization problem is formulated for simultaneously maximizing data rate and harvested energy. To tackle this problem, a two-stage algorithm is proposed. Specifically, the successive convex approximation (SCA) method is first invoked for minimizing the large-scale path loss, which is followed by the fine-tuning method for the phase alignment. 2) For the multiple IUs/EUs scenario, three multiple access schemes are considered, i.e., frequency division multiple access (FDMA), time division multiple access (TDMA), and non-orthogonal multiple access (NOMA). The corresponding multi-objective optimization problem (MOOP) that simultaneously maximizes the minimum data rate and minimum harvested energy is formulated for ensuring users' fairness. To address this problem, we adopt the $ε$-constraint method to first convert the intractable MOOPs to single-objective optimization problems (SOOPs). Then, for the SOOP under each multiple access protocol, the particle swarm optimization (PSO) and convex optimization methods are adopted for solving the pinching beamforming and resource allocation problems, respectively. Simulation results unveil that: i) PASS can achieve a significantly superior rate-energy region compared to conventional fixed-position antenna systems for pinching beamforming; and ii) by exploiting the time-switching feature, TDMA can outperform both NOMA and FDMA for the multiple IUs/EUs scenario.

eess.SP↗

arXiv2025

Flash-DMD: Towards High-Fidelity Few-Step Image Generation with Efficient Distillation and Joint Reinforcement Learning

Diffusion Models have emerged as a leading class of generative models, yet their iterative sampling process remains computationally expensive. Timestep distillation is a promising technique to accelerate generation, but it often requires extensive training and leads to image quality degradation. Furthermore, fine-tuning these distilled models for specific objectives, such as aesthetic appeal or user preference, using Reinforcement Learning (RL) is notoriously unstable and easily falls into reward hacking. In this work, we introduce Flash-DMD, a novel framework that enables fast convergence with distillation and joint RL-based refinement. Specifically, we first propose an efficient timestep-aware distillation strategy that significantly reduces training cost with enhanced realism, outperforming DMD2 with only $2.1\%$ its training cost. Second, we introduce a joint training scheme where the model is fine-tuned with an RL objective while the timestep distillation training continues simultaneously. We demonstrate that the stable, well-defined loss from the ongoing distillation acts as a powerful regularizer, effectively stabilizing the RL training process and preventing policy collapse. Extensive experiments on score-based and flow matching models show that our proposed Flash-DMD not only converges significantly faster but also achieves state-of-the-art generation quality in the few-step sampling regime, outperforming existing methods in visual quality, human preference, and text-image alignment metrics. Our work presents an effective paradigm for training efficient, high-fidelity, and stable generative models. Codes are coming soon.

cs.CV↗

arXiv2025

Survey of Vision-Language-Action Models for Embodied Manipulation

Embodied intelligence systems, which enhance agent capabilities through continuous environment interactions, have garnered significant attention from both academia and industry. Vision-Language-Action models, inspired by advancements in large foundation models, serve as universal robotic control frameworks that substantially improve agent-environment interaction capabilities in embodied intelligence systems. This expansion has broadened application scenarios for embodied AI robots. This survey comprehensively reviews VLA models for embodied manipulation. Firstly, it chronicles the developmental trajectory of VLA architectures. Subsequently, we conduct a detailed analysis of current research across 5 critical dimensions: VLA model structures, training datasets, pre-training methods, post-training methods, and model evaluation. Finally, we synthesize key challenges in VLA development and real-world deployment, while outlining promising future research directions.

cs.RO↗

arXiv2025

Stacking-orientation and twist-angle control on integer and fractional Chern insulators in moiré rhombohedral graphene

Rhombohedral-stacked multilayer graphene aligned with hexagonal boron nitride has emerged as an excellent platform for investigating exotic quantum phenomena arising from the interplay between electron correlations and nontrivial topology. However, the microscopic mechanism governing the emergence of both the integer and fractional Chern insulator states in this system remains an open question. In this work, we systematically investigate the electrical transport properties of RMG/hBN moiré devices with controlled alignment orientations and twist angles. We demonstrate that alignment orientation strongly modulates correlated phenomena in the moiré-proximal regime, while having negligible influence on the formation of integer and fractional Chern insulators in the moiré-distant regime. Instead, the moiré periodicity, tuned by the twist angle, serves as the key parameter controlling the stability of these correlated topological states in the moiré-distant regime. Furthermore, in the moiré-proximal regime of one specific alignment, we observe anomalous Hall effect and a variety of competing phases near ν = 1, including integer Chern insulator states, extended Chern insulator states, and trivial insulators, whose stability is highly sensitive to both the applied displacement electric field and magnetic field. Our results underscore the critical role of stacking-alignment and twist-angle engineering in exploring novel quantum states based on rhombohedral-stacked multilayer graphene moiré systems.

cond-mat.mes-hall↗

arXiv2025

Ultrafast magnetic moment transfer and bandgap renormalization in monolayer FeCl$_2$

The microscopic origin of laser-induced ultrafast demagnetization remains an open question, to which the non-thermal electronic distribution plays a vital role at the initial stage. Herein, we investigate the connection between the non-thermal electronic distribution and the ultrafast spin dynamics as well as the electronic structure evolution in ferromagnetic FeCl$_2$ monolayer using real-time time-dependent density functional theory (rt-TDDFT) with self-consistent Hubbard $U$ correction. Our simulations reveal that femtosecond laser pulses induce ultrafast magnetic moment transfer from Fe to Cl atoms. More importantly, through a comprehensive analysis of orbital-resolved electronic structure, we elucidate the microscopic origin of this transfer, attributing it to specific intra-atomic and inter-atomic charge transfer pathways driven by non-thermal excitations. The extent of demagnetization of Fe atoms exhibits a non-monotonic dependence on the laser photon energy, reaching a maximum at the resonant excitation. In addition, the dynamical evolution of the band structure was studied based on the eigenstates of the instantaneous Hamiltonian. Under resonant excitation, the bandgap reduction reaches up to $41\%$ within tens of fs. These findings provide fundamental insights into ultrafast spin control and suggest a strategy to optically engineer the magnetism in two-dimensional magnetic materials.

cond-mat.mtrl-sci↗

arXiv2025

Dexbotic: Open-Source Vision-Language-Action Toolbox

In this paper, we present Dexbotic, an open-source Vision-Language-Action (VLA) model toolbox based on PyTorch. It aims to provide a one-stop VLA research service for professionals in the field of embodied intelligence. It offers a codebase that supports multiple mainstream VLA policies simultaneously, allowing users to reproduce various VLA methods with just a single environment setup. The toolbox is experiment-centric, where the users can quickly develop new VLA experiments by simply modifying the Exp script. Moreover, we provide much stronger pretrained models to achieve great performance improvements for state-of-the-art VLA policies. Dexbotic will continuously update to include more of the latest pre-trained foundation models and cutting-edge VLA models in the industry.

cs.RO↗

arXiv2025

SpecExit: Accelerating Large Reasoning Model via Speculative Exit

Despite their strong performance on reasoning tasks, large reasoning models (LRMs) often suffer from overthinking, producing unnecessarily long outputs and incurring high end-to-end latency, a significant limitation to their real-world deployment. To address overthinking, early-exit mechanisms have been proposed to terminate reasoning before typical completion, showing that this approach can effectively shorten generation length with minimal impact on accuracy. However, their reliance on probing mechanisms introduces a detection overhead that limits their end-to-end latency gains and compromises their generalizability across diverse problems. Inspired by the use of hidden states in speculative decoding, we propose SpecExit, a novel framework that predicts both future tokens and an early-exit signal directly from a lightweight draft model without probing overhead. Our method offers significant improvements, reducing average generation length by 66\% and achieving a 2.5x speedup in end-to-end latency compared to the speculative decoding baseline, without compromising accuracy. Our method leverages the inherent signals from hidden states to provide effective early-exit signals, suggesting broader use of hidden states for efficient reasoning. Our code is available at https://github.com/Tencent/AngelSlim.

cs.AI↗

arXiv2025

OSCAR: One-Step Diffusion Codec Across Multiple Bit-rates

Pretrained latent diffusion models have shown strong potential for lossy image compression, owing to their powerful generative priors. Most existing diffusion-based methods reconstruct images by iteratively denoising from random noise, guided by compressed latent representations. While these approaches have achieved high reconstruction quality, their multi-step sampling process incurs substantial computational overhead. Moreover, they typically require training separate models for different compression bit-rates, leading to significant training and storage costs. To address these challenges, we propose a one-step diffusion codec across multiple bit-rates. termed OSCAR. Specifically, our method views compressed latents as noisy variants of the original latents, where the level of distortion depends on the bit-rate. This perspective allows them to be modeled as intermediate states along a diffusion trajectory. By establishing a mapping from the compression bit-rate to a pseudo diffusion timestep, we condition a single generative model to support reconstructions at multiple bit-rates. Meanwhile, we argue that the compressed latents retain rich structural information, thereby making one-step denoising feasible. Thus, OSCAR replaces iterative sampling with a single denoising pass, significantly improving inference efficiency. Extensive experiments demonstrate that OSCAR achieves superior performance in both quantitative and visual quality metrics. The code and models are available at https://github.com/jp-guo/OSCAR.

eess.IV↗

arXiv2025

RoboChallenge: Large-scale Real-robot Evaluation of Embodied Policies

Testing on real machines is indispensable for robotic control algorithms. In the context of learning-based algorithms, especially VLA models, demand for large-scale evaluation, i.e. testing a large number of models on a large number of tasks, is becoming increasingly urgent. However, doing this right is highly non-trivial, especially when scalability and reproducibility is taken into account. In this report, we describe our methodology for constructing RoboChallenge, an online evaluation system to test robotic control algorithms, and our survey of recent state-of-the-art VLA models using our initial benchmark Table30.

cs.RO↗