arXiv subjects

Jie Wang

Publications and source records attributed to Jie Wang.

At least 181 records · Page 10Linked to original sources

arXiv2025

DVAGen: Dynamic Vocabulary Augmented Generation

Language models trained with a fixed vocabulary struggle to generalize to novel or out-of-vocabulary words, limiting their flexibility in handling diverse token combinations. Existing dynamic vocabulary approaches attempt to address this limitation but face challenges such as fragmented codebases, lack of support for modern LLMs, and limited inference scalability. To overcome these issues, we introduce DVAGen, a fully open-source, unified framework designed for training, evaluation, and visualization of dynamic vocabulary-augmented language models. Our framework modularizes the pipeline for ease of customization, integrates seamlessly with open-source LLMs, and is the first to provide both CLI and WebUI tools for real-time result inspection. We validate the effectiveness of dynamic vocabulary methods on modern LLMs and demonstrate support for batch inference, significantly improving inference throughput.

cs.CL↗

arXiv2025

Probabilistic Constellation Shaping for OFDM ISAC Signals Under Temporal-Frequency Filtering

Integrated sensing and communications (ISAC) is considered an innovative technology in sixth-generation (6G) wireless networks, where utilizing orthogonal frequency division multiplexing (OFDM) communication signals for sensing provides a cost-effective solution for implementing ISAC. However, the sensing performance of matched and mismatched filtering schemes can be significantly deteriorated due to the signaling randomness induced by finite-alphabet modulations with nonconstant modulus, such as quadrature amplitude modulation (QAM) constellations. Therefore, improving sensing performance without significantly compromising communication capability (i.e., maintaining randomness), remains a challenging task. To that end, we propose a unified probabilistic constellation shaping (PCS) framework that is compatible with both matched and mismatched filtering schemes, by maximizing the communication rate while imposing constraints on mean square error (MSE) of sensing channel state information (CSI), power, and probability distribution. Specifically, the MSE of sensing CSI is leveraged to optimize sensing capability, which is illustrated to be a more comprehensive metric compared to the output SNR after filtering (SNRout) and integrated sidelobes ratio (ISLR). Additionally, the internal relationships among these three sensing metrics are explicitly analyzed. Finally, both simulations and field measurements validate the efficiency of proposed PCS approach in achieving a flexible S&C trade-off, as well as its credibility in enhancing 6G wireless transmission in real-world scenarios.

eess.SP↗

arXiv2025

Logo-VGR: Visual Grounded Reasoning for Open-world Logo Recognition

Recent advances in multimodal large language models (MLLMs) have been primarily evaluated on general-purpose benchmarks, while their applications in domain-specific scenarios, such as intelligent product moderation, remain underexplored. To address this gap, we introduce an open-world logo recognition benchmark, a core challenge in product moderation. Unlike traditional logo recognition methods that rely on memorizing representations of tens of thousands of brands-an impractical approach in real-world settings-our proposed method, Logo-VGR, enables generalization to large-scale brand recognition with supervision from only a small subset of brands. Specifically, we reformulate logo recognition as a comparison-based task, requiring the model to match product images with candidate logos rather than directly generating brand labels. We further observe that existing models tend to overfit by memorizing brand distributions instead of learning robust multimodal reasoning, which results in poor performance on unseen brands. To overcome this limitation, Logo-VGR introduces a new paradigm of domain-specific multimodal reasoning: Logo Perception Grounding injects domain knowledge, and Logo-Guided Visual Grounded Reasoning enhances the model's reasoning capability. Experimental results show that Logo-VGR outperforms strong baselines by nearly 10 points in OOD settings, demonstrating superior generalization.

cs.CV↗

arXiv2025

TransRec: Learning Transferable Recommendation from Mixture-of-Modality Feedback

Learning large-scale pre-trained models on broad-ranging data and then transfer to a wide range of target tasks has become the de facto paradigm in many machine learning (ML) communities. Such big models are not only strong performers in practice but also offer a promising way to break out of the task-specific modeling restrictions, thereby enabling task-agnostic and unified ML systems. However, such a popular paradigm is mainly unexplored by the recommender systems (RS) community. A critical issue is that standard recommendation models are primarily built on categorical identity features. That is, the users and the interacted items are represented by their unique IDs, which are generally not shareable across different systems or platforms. To pursue the transferable recommendations, we propose studying pre-trained RS models in a novel scenario where a user's interaction feedback involves a mixture-of-modality (MoM) items, e.g., text and images. We then present TransRec, a very simple modification made on the popular ID-based RS framework. TransRec learns directly from the raw features of the MoM items in an end-to-end training manner and thus enables effective transfer learning under various scenarios without relying on overlapped users or items. We empirically study the transferring ability of TransRec across four different real-world recommendation settings. Besides, we look at its effects by scaling source and target data size. Our results suggest that learning neural recommendation models from MoM feedback provides a promising way to realize universal RS.

cs.IR↗

arXiv2025

Sequential-NIAH: A Needle-In-A-Haystack Benchmark for Extracting Sequential Needles from Long Contexts

Evaluating the ability of large language models (LLMs) to process lengthy contexts is critical, especially for retrieving query-relevant information embedded within them. We introduce Sequential-NIAH, a benchmark specifically designed to evaluate the capability of LLMs to extract sequential information items (known as \emph{needles}) from long contexts. The benchmark includes three needle generation pipelines: synthetic-temporal, real-temporal, and real-logical orders, with context lengths ranging from 8K to 128K, which comprises 14,000 samples (2,000 for testing). To facilitate the evaluation of this benchmark, we trained an evaluation model that assesses the correctness of LLM responses by comparing their completeness and sequential consistency against the ground truth, which provides a more reliable evaluation metric than GPT-4 or Claude. We conducted experiments on six well-known LLMs, revealing that even the best-performing model achieved a maximum accuracy of only 63.50% on test set of this benchmark. Further analysis highlights the growing challenges posed by increasing the context length or the number of needles, underscoring substantial room for improvement of LLMs. Additionally, noise analysis validates the reliability and challenge of the benchmark, making Sequential-NIAH an important reference for advancing research on long text information extraction capabilities of LLMs.

cs.CL↗

arXiv2025

Highly tunable Gilbert damping in two-dimensional van der Waals ferromagnet Fe3GaTe2: From bilayer to the twisted bilayer

Van der Waals ferromagnet Fe3GaTe2 possesses both a high Curie temperature and robust perpendicular magnetic anisotropy, holding promise for practical spintronic applications. In particular, understanding and engineering its Gilbert damping which determines magnetization dynamics are crucial for its applications. Here, we investigate the Gilbert damping of bilayer and the twisted bilayer Fe3GaTe2 through first-principles calculations. For the bilayer Fe3GaTe2, we obtain a quite low Gilbert damping when its magnetization is along the z axis at room temperature. In addition, the bilayer Fe3GaTe2 exhibits a large orientational anisotropy of Gilbert damping when its magnetization is rotated from the magnetic easy axis to the hard one. Such anisotropy is attributed to the distinct band structures caused by the anisotropic spin-orbit coupling. Surprisingly, we find that twisting the bilayer Fe3GaTe2 can effectively reduce the Gilbert damping for the perpendicular magnetization, and enhance the orientational anisotropy of Gilbert damping up to 635% when rotating the magnetization from the magnetic easy axis to the hard one. These findings open up an entirely new avenue for the manipulation of Gilbert damping and its anisotropy in two-dimensional van der Waals ferromagnets.

cond-mat.mtrl-sci↗

arXiv2025

Nonlinear reconstruction of general dark energy theories

The large variety and number of dark energy (DE) theories make it impractical to perform detailed analyses on a case-by-case basis, which has motivated proposals to ``parameterize" theories to reduce the size of theory space. The leading approach to do this is the effective field theory of dark energy (EFTofDE), which can describe general Horndeski-type theories with a small number of observationally accessible time-dependent functions. However, the EFTofDE primarily works for linear perturbations, and extending it to obtain a fully non-linear description of DE theories, which is critical for theories with screening mechanisms, is challenging. In this paper, we present a general method for reconstructing the non-linear DE Lagrangian from the background expansion history and certain linear-perturbation quantities, building upon the EFTofDE framework. Using numerical examples, we demonstrate that this method is applicable to a wide range of single-scalar-field dark energy and modified gravity theories, including quintessence, scalar-tensor theory, $k$-essence, and generalized cubic Galileon with shift symmetry. For each of these theories, we discuss the validity of the method and factors affecting its results. While this method involves solving differential equations, we find that the initial conditions are not important for quintessence, scalar-tensor theory and $k$-essence, while for shift-symmetric cubic Galileon, the generic tracker solution can help transform differential equations into algebraic equations. This offers a useful framework to connect cosmological observations at the background and linear-perturbation levels to the underlying non-linear dynamics of dark energy, and will enable cosmological simulations to analyze and examine DE theories systematically and in much greater detail.

astro-ph.CO↗

arXiv2025

Deep Learning for Personalized Binaural Audio Reproduction

Personalized binaural audio reproduction is the basis of realistic spatial localization, sound externalization, and immersive listening, directly shaping user experience and listening effort. This survey reviews recent advances in deep learning for this task and organizes them by generation mechanism into two paradigms: explicit personalized filtering and end-to-end rendering. Explicit methods predict personalized head-related transfer functions (HRTFs) from sparse measurements, morphological features, or environmental cues, and then use them in the conventional rendering pipeline. End-to-end methods map source signals directly to binaural signals, aided by other inputs such as visual, textual, or parametric guidance, and they learn personalization within the model. We also summarize the field's main datasets and evaluation metrics to support fair and repeatable comparison. Finally, we conclude with a discussion of key applications enabled by these technologies, current technical limitations, and potential research directions for deep learning-based spatial audio systems.

eess.AS↗

arXiv2025

BudgetThinker: Empowering Budget-aware LLM Reasoning with Control Tokens

Recent advancements in Large Language Models (LLMs) have leveraged increased test-time computation to enhance reasoning capabilities, a strategy that, while effective, incurs significant latency and resource costs, limiting their applicability in real-world time-constrained or cost-sensitive scenarios. This paper introduces BudgetThinker, a novel framework designed to empower LLMs with budget-aware reasoning, enabling precise control over the length of their thought processes. We propose a methodology that periodically inserts special control tokens during inference to continuously inform the model of its remaining token budget. This approach is coupled with a comprehensive two-stage training pipeline, beginning with Supervised Fine-Tuning (SFT) to familiarize the model with budget constraints, followed by a curriculum-based Reinforcement Learning (RL) phase that utilizes a length-aware reward function to optimize for both accuracy and budget adherence. We demonstrate that BudgetThinker significantly surpasses strong baselines in maintaining performance across a variety of reasoning budgets on challenging mathematical benchmarks. Our method provides a scalable and effective solution for developing efficient and controllable LLM reasoning, making advanced models more practical for deployment in resource-constrained and real-time environments.

cs.LG↗

arXiv2025

Giant Anomalous Hall Conductivity and Gilbert Damping in Room-temperature Ferromagnetic Half-Heusler Alloys PtMnBi

Half-Heusler alloys have emerged as promising candidates for novel spintronic applications due to their exceptional properties including the high Curie temperature (TC) above room temperature and large anomalous Hall conductivity (AHC). In this work, we systematically study the magnetic and electronic properties of PtMnBi in α-, \{beta}-, and γ-phase using first-principles calculations and Monte Carlo simulations. The three phases are found to be ferromagnetic metals. In particular, the α-phase PtMnBi shows a high TC up to 802 K and a relatively large Gilbert damping of 0.085. Additionally, the γ-phase PtMnBi possesses a non-negligible AHC, reaching 203 Ω-1cm-1 at the Fermi level. To evaluate its potential in nanoscale devices, we further investigate the α-phase PtMnBi thin films. The Gilbert dampings of α-phase PtMnBi thin films varies with film thickness and we attribute this variation to the distinct band structures at the high-symmetry point Γ, which arise from differences in film thickness. Moreover, the 1-layer (1L) α-phase thin film retains robust ferromagnetism (TC = 688 K) and shows enhanced Gilbert damping (0.14) and AHC (1116 Ω-1cm-1) compared to the bulk. Intriguingly, under a 2% in-plane biaxial compressive strain, the Gilbert damping of 1L α-phase PtMnBi thin film increases to 0.17 and the AHC reaches 2386 Ω-1cm-1. The coexistence of giant Gilbert damping and large AHC makes α-phase PtMnBi a compelling platform for practical spintronic applications, and highlights the potential of half-Heusler alloys in spintronic device design.

cond-mat.mtrl-sci↗

arXiv2025

CURE: Critical-Token-Guided Re-Concatenation for Entropy-Collapse Prevention

Recent advances in Reinforcement Learning with Verified Reward (RLVR) have driven the emergence of more sophisticated cognitive behaviors in large language models (LLMs), thereby enhancing their reasoning capabilities. However, in prior RLVR pipelines, the repeated use of static initial-state sampling drawn exactly from the dataset distribution during each sampling phase produced overly deterministic, low diversity model behavior, which manifested as rapid entropy collapse and hindered sustained performance gains during prolonged training. To address this issue, we introduce CURE (Critical-token-gUided Re concatenation for Entropy-collapse prevention), a two-stage framework that balances exploration and exploitation. Specifically, in the first stage, to deliberately steer the model toward novel yet coherent contexts, we re-generate at high-entropy critical tokens and jointly optimize the original and the branched trajectories. The further comparison with vanilla DAPO shows that the regeneration process achieves a better performance on math reasoning tasks while sustaining a high-level entropy degree for exploration. In the second stage, we continue training with static initial-state sampling by DAPO, intentionally placing the model in a familiar state to gradually strengthen exploitation. Extensive experiments on Qwen-2.5-Math-7B show that, compared to other RLVR methods, CURE achieves a 5% performance gain across six math benchmarks, establishing state-of-the-art performance in both entropy and accuracy. A series of experiments further validate the effectiveness of our approach. Code is available at https://github.com/bytedance/CURE.

cs.LG↗

arXiv2025

MeshCoder: LLM-Powered Structured Mesh Code Generation from Point Clouds

Reconstructing 3D objects into editable programs is pivotal for applications like reverse engineering and shape editing. However, existing methods often rely on limited domain-specific languages (DSLs) and small-scale datasets, restricting their ability to model complex geometries and structures. To address these challenges, we introduce MeshCoder, a novel framework that reconstructs complex 3D objects from point clouds into editable Blender Python scripts. We develop a comprehensive set of expressive Blender Python APIs capable of synthesizing intricate geometries. Leveraging these APIs, we construct a large-scale paired object-code dataset, where the code for each object is decomposed into distinct semantic parts. Subsequently, we train a multimodal large language model (LLM) that translates 3D point cloud into executable Blender Python scripts. Our approach not only achieves superior performance in shape-to-code reconstruction tasks but also facilitates intuitive geometric and topological editing through convenient code modifications. Furthermore, our code-based representation enhances the reasoning capabilities of LLMs in 3D shape understanding tasks. Together, these contributions establish MeshCoder as a powerful and flexible solution for programmatic 3D shape reconstruction and understanding. The project homepage is available at \href{https://daibingquan.github.io/MeshCoder}{this link}.

cs.GR↗

arXiv2025

We-Math 2.0: A Versatile MathBook System for Incentivizing Visual Mathematical Reasoning

Multimodal Large Language Models (MLLMs) have demonstrated impressive capabilities across various tasks, but still struggle with complex mathematical reasoning. Existing research primarily focuses on dataset construction and method optimization, often overlooking two critical aspects: comprehensive knowledge-driven design and model-centric data space modeling. In this paper, we introduce We-Math 2.0, a unified system that integrates a structured mathematical knowledge system, model-centric data space modeling, and a reinforcement learning (RL)-based training paradigm to comprehensively enhance the mathematical reasoning abilities of MLLMs. The key contributions of We-Math 2.0 are fourfold: (1) MathBook Knowledge System: We construct a five-level hierarchical system encompassing 491 knowledge points and 1,819 fundamental principles. (2) MathBook-Standard & Pro: We develop MathBook-Standard, a dataset that ensures broad conceptual coverage and flexibility through dual expansion. Additionally, we define a three-dimensional difficulty space and generate 7 progressive variants per problem to build MathBook-Pro, a challenging dataset for robust training. (3) MathBook-RL: We propose a two-stage RL framework comprising: (i) Cold-Start Fine-tuning, which aligns the model with knowledge-oriented chain-of-thought reasoning; and (ii) Progressive Alignment RL, leveraging average-reward learning and dynamic data scheduling to achieve progressive alignment across difficulty levels. (4) MathBookEval: We introduce a comprehensive benchmark covering all 491 knowledge points with diverse reasoning step distributions. Experimental results show that MathBook-RL performs competitively with existing baselines on four widely-used benchmarks and achieves strong results on MathBookEval, suggesting promising generalization in mathematical reasoning.

cs.AI↗

arXiv2025

A First Look at Predictability and Explainability of Pre-request Passenger Waiting Time in Ridesharing Systems

Passenger waiting time prediction plays a critical role in enhancing both ridesharing user experience and platform efficiency. While most existing research focuses on post-request waiting time prediction with knowing the matched driver information, pre-request waiting time prediction (i.e., before submitting a ride request and without matching a driver) is also important, as it enables passengers to plan their trips more effectively and enhance the experience of both passengers and drivers. However, it has not been fully studied by existing works. In this paper, we take the first step toward understanding the predictability and explainability of pre-request passenger waiting time in ridesharing systems. Particularly, we conduct an in-depth data-driven study to investigate the impact of demand&supply dynamics on passenger waiting time. Based on this analysis and feature engineering, we propose FiXGBoost, a novel feature interaction-based XGBoost model designed to predict waiting time without knowing the assigned driver information. We further perform an importance analysis to quantify the contribution of each factor. Experiments on a large-scale real-world ridesharing dataset including over 30 million trip records show that our FiXGBoost can achieve a good performance for pre-request passenger waiting time prediction with high explainability.

cs.AI↗

arXiv2025

A Trustworthy Method for Multimodal Emotion Recognition

Existing emotion recognition methods mainly focus on enhancing performance by employing complex deep models, typically resulting in significantly higher model complexity. Although effective, it is also crucial to ensure the reliability of the final decision, especially for noisy, corrupted and out-of-distribution data. To this end, we propose a novel emotion recognition method called trusted emotion recognition (TER), which utilizes uncertainty estimation to calculate the confidence value of predictions. TER combines the results from multiple modalities based on their confidence values to output the trusted predictions. We also provide a new evaluation criterion to assess the reliability of predictions. Specifically, we incorporate trusted precision and trusted recall to determine the trusted threshold and formulate the trusted Acc. and trusted F1 score to evaluate the model's trusted performance. The proposed framework combines the confidence module that accordingly endows the model with reliability and robustness against possible noise or corruption. The extensive experimental results validate the effectiveness of our proposed model. The TER achieves state-of-the-art performance on the Music-video, achieving 82.40% Acc. In terms of trusted performance, TER outperforms other methods on the IEMOCAP and Music-video, achieving trusted F1 scores of 0.7511 and 0.9035, respectively.

cs.CV↗

arXiv2025

The Poisson noise in modeling the redshift-space distortion at large scales

We investigate the errors in modeling the redshift-space distortion (RSD) effect at large linear scales, using data from the Millennium simulation. While standard theoretical templates, such as the Kaiser formula and the TNS method, could precisely model RSD for individual large-scale modes, we find that for tracers with number densities lower than $\sim10^{-3}({\rm Mpc}/h)^{-3}$, there is a few-percent level bias in the predicted power spectrum. This error arises due to the amplification of intrinsic Poisson noise during RSD modeling from real-space power spectrum. This amplified noise can be analytically expressed as $1 + ε/[{\bar{n}P}({1+ε})]$, with $ε=2β/3+β^2/5$, where $P$ denotes the real-space tracer power spectrum and $β\equiv f/b$. Specifically, for halos with a number density of around $5\times10^{-4}({\rm Mpc}/h)^{-3}$, this phenomenon results in an additional systematic error of 2.5\%. Our result suggests that caution is necessary when directly modeling redshift-space distortions (RSD) using real-space power spectra of tracers obtained from simulations or actual surveys. This caution is particularly pertinent in scenarios where emulators trained on simulation data forecast the real-space tracer power spectrum, as well as in baryon acoustic oscillation (BAO) reconstruction using galaxy samples, for which we estimate that shot noise could introduce random errors of about one-third in the displacement field, potentially diminishing the effectiveness of the BAO peak sharpening.

astro-ph.CO↗

arXiv2025

A Structure-aware and Motion-adaptive Framework for 3D Human Pose Estimation with Mamba

Recent Mamba-based methods for the pose-lifting task tend to model joint dependencies by 2D-to-1D mapping with diverse scanning strategies. Though effective, they struggle to model intricate joint connections and uniformly process all joint motion trajectories while neglecting the intrinsic differences across motion characteristics. In this work, we propose a structure-aware and motion-adaptive framework to capture spatial joint topology along with diverse motion dynamics independently, named as SAMA. Specifically, SAMA consists of a Structure-aware State Integrator (SSI) and a Motion-adaptive State Modulator (MSM). The Structure-aware State Integrator is tasked with leveraging dynamic joint relationships to fuse information at both the joint feature and state levels in the state space, based on pose topology rather than sequential state transitions. The Motion-adaptive State Modulator is responsible for joint-specific motion characteristics recognition, thus applying tailored adjustments to diverse motion patterns across different joints. Through the above key modules, our algorithm enables structure-aware and motion-adaptive pose lifting. Extensive experiments across multiple benchmarks demonstrate that our algorithm achieves advanced results with fewer computational costs.

cs.CV↗

arXiv2025

Dynamic mapping from static labels: remote sensing dynamic sample generation with temporal-spectral embedding

Accurate remote sensing geographic mapping requires timely and representative samples. However, rapid land surface changes often render static samples obsolete within months, making manual sample updates labor-intensive and unsustainable. To address this challenge, we propose TasGen, a two-stage Temporal spectral-aware Automatic Sample Generation method for generating dynamic training samples from single-date static labels without human intervention. Land surface dynamics often manifest as anomalies in temporal-spectral sequences. %These anomalies are multivariate yet unified: temporal, spectral, or joint anomalies stem from different mechanisms and cannot be naively coupled, as this may obscure the nature of changes. Yet, any land surface state corresponds to a coherent temporal-spectral signature, which would be lost if the two dimensions are modeled separately. To effectively capture these dynamics, TasGen first disentangles temporal and spectral features to isolate their individual contributions, and then couples them to model their synergistic interactions. In the first stage, we introduce a hierarchical temporal-spectral variational autoencoder (HTS-VAE) with a dual-dimension embedding to learn low-dimensional latent patterns of normal samples by first disentangling and then jointly embedding temporal and spectral information. This temporal-spectral embedding enables robust anomaly detection by identifying deviations from learned joint patterns. In the second stage, a classifier trained on stable samples relabels change points across time to generate dynamic samples. To not only detect but also explain surface dynamics, we further propose an anomaly interpretation method based on Gibbs sampling, which attributes changes to specific spectral-temporal dimensions.

eess.IV↗