arXiv subjects

Dong Wang

Publications and source records attributed to Dong Wang.

At least 235 records · Page 13Linked to original sources

arXiv2025

Algorithmic Complexity Attacks on All Learned Cardinality Estimators: A Data-centric Approach

Learned cardinality estimators show promise in query cardinality prediction, yet they universally exhibit fragility to training data drifts, posing risks for real-world deployment. This work is the first to theoretical investigate how minimal data-level drifts can maximally degrade the accuracy of learned estimators. We propose data-centric algorithmic complexity attacks against learned estimators in a black-box setting, proving that finding the optimal attack strategy is NP-Hard. To address this, we design a polynomial-time approximation algorithm with a $(1-κ)$ approximation ratio. Extensive experiments demonstrate our attack's effectiveness: on STATS-CEB and IMDB-JOB benchmarks, modifying just 0.8\% of training tuples increases the 90th percentile Qerror by three orders of magnitude and raises end-to-end processing time by up to 20$\times$. Our work not only reveals critical vulnerabilities in deployed learned estimators but also provides the first unified worst-case theoretical analysis of their fragility under data updates. Additionally, we identify two countermeasures to mitigate such black-box attacks, offering insights for developing robust learned database optimizers.

cs.DB↗

arXiv2025

Spectral Bounds on Entropy and Ergotropy via Statistical Effective Temperature in Classical Polarization and Quantum Thermal States

We formulate a unified definition of the statistical effective temperature (SET) for finite-dimensional classical and quantum systems using dimension-dependent indices of purity derived from the eigenvalue spectrum. This spectral approach bypasses the need for Hamiltonians or energy gaps and remains applicable to both quantum density matrices and classical polarization coherency matrices. The SET framework naturally describes the divergence of inverse temperature near pure, non-degenerate states, consistent with the third law. Using entropy-SET diagrams, we explore spectral bounds in two-, three-, and four-level systems, which reveal physically realizable entropy regions, rank-dependent constraints, and cusp-like features. A Hamiltonian-free parametric spectrum ansatz provides a universal reference curve within these bounds. Furthermore, we derive spectral bounds on ergotropy as a function of entropy and SET, which quantify the maximum extractable work under passive constraints and introduce the notion of structured-states, engineered spectral configurations that saturate these bounds. Our analysis shows that SET serves as a thermodynamically meaningful and operationally relevant quantity for bounding entropy and ergotropy in both classical polarization systems and quantum thermal states.

quant-ph↗

arXiv2025

Hume: Introducing System-2 Thinking in Visual-Language-Action Model

Humans practice slow thinking before performing actual actions when handling complex tasks in the physical world. This thinking paradigm, recently, has achieved remarkable advancement in boosting Large Language Models (LLMs) to solve complex tasks in digital domains. However, the potential of slow thinking remains largely unexplored for robotic foundation models interacting with the physical world. In this work, we propose Hume: a dual-system Vision-Language-Action (VLA) model with value-guided System-2 thinking and cascaded action denoising, exploring human-like thinking capabilities of Vision-Language-Action models for dexterous robot control. System 2 of Hume implements value-Guided thinking by extending a Vision-Language-Action Model backbone with a novel value-query head to estimate the state-action value of predicted actions. The value-guided thinking is conducted by repeat sampling multiple action candidates and selecting one according to state-action value. System 1 of Hume is a lightweight reactive visuomotor policy that takes System 2 selected action and performs cascaded action denoising for dexterous robot control. At deployment time, System 2 performs value-guided thinking at a low frequency while System 1 asynchronously receives the System 2 selected action candidate and predicts fluid actions in real time. We show that Hume outperforms the existing state-of-the-art Vision-Language-Action models across multiple simulation benchmark and real-robot deployments.

cs.RO↗

arXiv2025

Hunyuan-TurboS: Advancing Large Language Models through Mamba-Transformer Synergy and Adaptive Chain-of-Thought

As Large Language Models (LLMs) rapidly advance, we introduce Hunyuan-TurboS, a novel large hybrid Transformer-Mamba Mixture of Experts (MoE) model. It synergistically combines Mamba's long-sequence processing efficiency with Transformer's superior contextual understanding. Hunyuan-TurboS features an adaptive long-short chain-of-thought (CoT) mechanism, dynamically switching between rapid responses for simple queries and deep "thinking" modes for complex problems, optimizing computational resources. Architecturally, this 56B activated (560B total) parameter model employs 128 layers (Mamba2, Attention, FFN) with an innovative AMF/MF block pattern. Faster Mamba2 ensures linear complexity, Grouped-Query Attention minimizes KV cache, and FFNs use an MoE structure. Pre-trained on 16T high-quality tokens, it supports a 256K context length and is the first industry-deployed large-scale Mamba model. Our comprehensive post-training strategy enhances capabilities via Supervised Fine-Tuning (3M instructions), a novel Adaptive Long-short CoT Fusion method, Multi-round Deliberation Learning for iterative improvement, and a two-stage Large-scale Reinforcement Learning process targeting STEM and general instruction-following. Evaluations show strong performance: overall top 7 rank on LMSYS Chatbot Arena with a score of 1356, outperforming leading models like Gemini-2.0-Flash-001 (1352) and o4-mini-2025-04-16 (1345). TurboS also achieves an average of 77.9% across 23 automated benchmarks. Hunyuan-TurboS balances high performance and efficiency, offering substantial capabilities at lower inference costs than many reasoning models, establishing a new paradigm for efficient large-scale pre-trained models.

cs.CL↗

arXiv2025

Neural Scoring: A Refreshed End-to-End Approach for Speaker Recognition in Complex Conditions

Modern speaker verification systems primarily rely on speaker embeddings, followed by verification based on cosine similarity between the embedding vectors of the enrollment and test utterances. While effective, these methods struggle with multi-talker speech due to the unidentifiability of embedding vectors. In this paper, we propose Neural Scoring (NS), a refreshed end-to-end framework that directly estimates verification posterior probabilities without relying on test-side embeddings, making it more robust to complex conditions, e.g., with multiple talkers. To make the training of such an end-to-end model more efficient, we introduce a large-scale trial e2e training (LtE2E) strategy, where each test utterance pairs with a set of enrolled speakers, thus enabling the processing of large-scale verification trials per batch. Experiments on the VoxCeleb dataset demonstrate that NS consistently outperforms both the baseline and competitive methods across various conditions, achieving an overall 70.36% reduction in EER compared to the baseline.

cs.SD↗

arXiv2025

Large-eddy simulation of turbulent spray flames: Effects of scalar correlation and enthalpy reduction in flamelet modeling

Numerical modeling of turbulent spray combustion provides a promising tool for advanced engine design. In spray flames, the droplet evaporation not only reduces the ambient gas temperature, but also influences flame structure by generating substantial local fluctuations of the mixture fraction $\widetilde{Z}$ and progress variable $\widetilde{c}$. These two scalars, conventionally assumed independent in flamelet models, exhibit significant correlations arising from the coupling among evaporation, turbulent mixing and chemical reactions. This study proposes a six-dimensional flamelet-generated manifolds (FGM) method, considering the evaporation-induced specific enthalpy reduction and scalar correlation. A novel joint presumed probability density function (PDF) method is derived using the copula theory, achieving rapid grid convergence and good feasibility. Large-eddy simulation (LES) is performed on the Sydney ethanol turbulent spray flames (EtF1, EtF4 and EtF7), which feature different ethanol mass flow rates and jet Reynolds numbers. Both gas and liquid phase statistics show good agreement with experimental data across the three flames. The incorporation of specific enthalpy reduction and scalar correlation in FGM modeling improves gas temperature predictions, along with enhanced liquid-phase prediction through refined gas-field resolution. The correlation coefficient of $\widetilde{Z}$ and $\widetilde{c}$ is found to be a competing result of local evaporation and combustion, since evaporation elevates $\widetilde{Z}$ and dilutes reaction products, whereas chemical reactions enhance $\widetilde{c}$ fluctuations.

physics.flu-dyn↗

arXiv2025

NaturalThoughts: Selecting and Distilling Reasoning Traces for General Reasoning Tasks

Recent work has shown that distilling reasoning traces from a larger teacher model via supervised finetuning outperforms reinforcement learning with the smaller student model alone (Guo et al. 2025). However, there has not been a systematic study of what kind of reasoning demonstrations from the teacher are most effective in improving the student model's reasoning capabilities. In this work we curate high-quality "NaturalThoughts" by selecting reasoning traces from a strong teacher model based on a large pool of questions from NaturalReasoning (Yuan et al. 2025). We first conduct a systematic analysis of factors that affect distilling reasoning capabilities, in terms of sample efficiency and scalability for general reasoning tasks. We observe that simply scaling up data size with random sampling is a strong baseline with steady performance gains. Further, we find that selecting difficult examples that require more diverse reasoning strategies is more sample-efficient to transfer the teacher model's reasoning skills. Evaluated on both Llama and Qwen models, training with NaturalThoughts outperforms existing reasoning datasets such as OpenThoughts, LIMO, etc. on general STEM reasoning benchmarks including GPQA-Diamond, MMLU-Pro and SuperGPQA.

cs.CL↗

arXiv2025

Strong and noise-tolerant entanglement in dissipative optomechanics

Macroscopic entanglement, as a critical quantum resource in quantum information science, has been extensively studied in coherent optomechanics over the past decades. However, entanglement in dissipative optomechanics, where the cavity linewidth depends on the position of the mechanical resonator, remains largely unexplored. In this work, we investigate quantum entanglement in a dissipative optomechanical system realized by a Michelson-Sagnac interferometer with a movable membrane. This configuration enables the switching between coherent and dissipative optomechanical couplings at will. With experimentally feasible parameters, we demonstrate that the steady-state mechanical displacement exhibits a nonlinear (linear) dependence on the driving power under coherent (dissipative) coupling. Furthermore, we show that the quantum entanglement generated via dissipative coupling is significantly stronger and more robust to noise than that generated via coherent coupling. When both coherent and dissipative couplings are simultaneously present, the entanglement is weakened due to quantum interference. Our results indicate that dissipative optomechanical coupling can be a promising route for engineering strong and noise-resilient quantum entanglement.

quant-ph↗

arXiv2025

Simulation studies of a high-repetition-rate electron-driven surface muon beamline at SHINE

A high-repetition-rate pulsed muon source operating at approximately 50\,kHz holds the potential to improve the sensitivity of various particle physics and material science experiments involving muons. In this article, we propose utilizing the high-repetition-rate pulsed electron beam at the SHINE facility to generate a surface muon beam. Our simulation studies indicate that an 8\,GeV, 100\,pC charge pulsed electron beam impinging on a copper target can produce up to $2 \times 10^{3}$ muons per pulse. Beamline optimization results demonstrate that approximately 60 surface muons per electron bunch can be efficiently transported to the end of the beamline. This translates to a surface muon rate of $3 \times 10^{6}\,μ^{+}$/s when the pulsed electron beam is operated at 50\,kHz, which is comparable to existing muon facilities. This high-repetition-rate pulsed muon beam, with its ideal time structure, represents a unique and pioneering effort once constructed. It serves as a model for building cost-effective muon sources at existing electron machines with GeV electron energies. In addition to the typical challenges encountered in conventional muon beamlines, such as the installation and construction of the target station and beamline, the removal of substantial quantities of positrons is also a major challenge. A potential solution to this issue is also discussed.

physics.acc-ph↗

arXiv2025

Exploiting Lightweight Hierarchical ViT and Dynamic Framework for Efficient Visual Tracking

Transformer-based visual trackers have demonstrated significant advancements due to their powerful modeling capabilities. However, their practicality is limited on resource-constrained devices because of their slow processing speeds. To address this challenge, we present HiT, a novel family of efficient tracking models that achieve high performance while maintaining fast operation across various devices. The core innovation of HiT lies in its Bridge Module, which connects lightweight transformers to the tracking framework, enhancing feature representation quality. Additionally, we introduce a dual-image position encoding approach to effectively encode spatial information. HiT achieves an impressive speed of 61 frames per second (fps) on the NVIDIA Jetson AGX platform, alongside a competitive AUC of 64.6% on the LaSOT benchmark, outperforming all previous efficient trackers.Building on HiT, we propose DyHiT, an efficient dynamic tracker that flexibly adapts to scene complexity by selecting routes with varying computational requirements. DyHiT uses search area features extracted by the backbone network and inputs them into an efficient dynamic router to classify tracking scenarios. Based on the classification, DyHiT applies a divide-and-conquer strategy, selecting appropriate routes to achieve a superior trade-off between accuracy and speed. The fastest version of DyHiT achieves 111 fps on NVIDIA Jetson AGX while maintaining an AUC of 62.4% on LaSOT.Furthermore, we introduce a training-free acceleration method based on the dynamic routing architecture of DyHiT. This method significantly improves the execution speed of various high-performance trackers without sacrificing accuracy. For instance, our acceleration method enables the state-of-the-art tracker SeqTrack-B256 to achieve a 2.68 times speedup on an NVIDIA GeForce RTX 2080 Ti GPU while maintaining the same AUC of 69.9% on the LaSOT.

cs.CV↗

arXiv2025

Cross from Left to Right Brain: Adaptive Text Dreamer for Vision-and-Language Navigation

Vision-and-Language Navigation (VLN) requires the agent to navigate by following natural instructions under partial observability, making it difficult to align perception with language. Recent methods mitigate this by imagining future scenes, yet they rely on vision-based synthesis, leading to high computational cost and redundant details. To this end, we propose to adaptively imagine key environmental semantics via \textit{language} form, enabling a more reliable and efficient strategy. Specifically, we introduce a novel Adaptive Text Dreamer (ATD), a dual-branch self-guided imagination policy built upon a large language model (LLM). ATD is designed with a human-like left-right brain architecture, where the left brain focuses on logical integration, and the right brain is responsible for imaginative prediction of future scenes. To achieve this, we fine-tune only the Q-former within both brains to efficiently activate domain-specific knowledge in the LLM, enabling dynamic updates of logical reasoning and imagination during navigation. Furthermore, we introduce a cross-interaction mechanism to regularize the imagined outputs and inject them into a navigation expert module, allowing ATD to jointly exploit both the reasoning capacity of the LLM and the expertise of the navigation model. We conduct extensive experiments on the R2R benchmark, where ATD achieves state-of-the-art performance with fewer parameters. The code is \href{https://github.com/zhangpingrui/Adaptive-Text-Dreamer}{here}.

cs.CV↗

arXiv2025

How Breakable Is Privacy: Probing and Resisting Model Inversion Attacks in Collaborative Inference

Collaborative inference (CI) improves computational efficiency for edge devices by transmitting intermediate features to cloud models. However, this process inevitably exposes feature representations to model inversion attacks (MIAs), enabling unauthorized data reconstruction. Despite extensive research, there is no established criterion for assessing the difficulty of MIA implementation, leaving a fundamental question unanswered: \textit{What factors truly and verifiably determine the attack's success in CI?} Moreover, existing defenses lack the theoretical foundation described above, making it challenging to regulate feature information effectively while ensuring privacy and minimizing computational overhead. These shortcomings introduce three key challenges: theoretical gap, methodological limitation, and practical constraint. To overcome these challenges, we propose the first theoretical criterion to assess MIA difficulty in CI, identifying mutual information, entropy, and effective information volume as key influencing factors. The validity of this criterion is demonstrated by using the mutual information neural estimator. Building on this insight, we propose SiftFunnel, a privacy-preserving framework to resist MIA while maintaining usability. Specifically, we incorporate linear and non-linear correlation constraints alongside label smoothing to suppress redundant information transmission, effectively balancing privacy and usability. To enhance deployability, the edge model adopts a funnel-shaped structure with attention mechanisms, strengthening privacy while reducing computational and storage burdens. Experiments show that, compared to state-of-the-art defense, SiftFunnel increases reconstruction error by $\sim$30\%, lowers mutual and effective information metrics by $\geq$50\%, and reduces edge burdens by almost $20\times$, while maintaining comparable usability.

cs.CR↗

arXiv2025

Entanglement and entropy uncertainty in black hole quantum atmosphere

In this work, we investigate the properties of Hawking radiation induced by the quantum atmosphere beyond the event horizon, by considering two detectors in Schwarzschild spacetime with the parameterized Hartle-Hawking temperature. \textcolor{black}{We explicitly study the dynamics of quantum entanglement and found that its characteristics are closely correlated with Hawking quantum radiation beyond the event horizon. Namely, its minimal value corresponds to the peak of Hawking radiation.} By virtue of the mutual information, we demonstrate the complementary relationship of the information distribution in the black hole. In addition, we detailedly discuss the influence of distance from the center of black hole to particle, radius of event horizon and Hartle-Hawking constant on the entropy uncertainty in the current scenario, and the results interestingly show that there exists an opposite correlation between the entanglement and the entropy uncertainty. It is believed that our observation could provide a new perspective for understanding the black hole information paradox and black hole thermodynamics.

gr-qc↗

arXiv2025

Integer Binary-Range Alignment Neuron for Spiking Neural Networks

Spiking Neural Networks (SNNs) are noted for their brain-like computation and energy efficiency, but their performance lags behind Artificial Neural Networks (ANNs) in tasks like image classification and object detection due to the limited representational capacity. To address this, we propose a novel spiking neuron, Integer Binary-Range Alignment Leaky Integrate-and-Fire to exponentially expand the information expression capacity of spiking neurons with only a slight energy increase. This is achieved through Integer Binary Leaky Integrate-and-Fire and range alignment strategy. The Integer Binary Leaky Integrate-and-Fire allows integer value activation during training and maintains spike-driven dynamics with binary conversion expands virtual timesteps during inference. The range alignment strategy is designed to solve the spike activation limitation problem where neurons fail to activate high integer values. Experiments show our method outperforms previous SNNs, achieving 74.19% accuracy on ImageNet and 66.2% mAP@50 and 49.1% mAP@50:95 on COCO, surpassing previous bests with the same architecture by +3.45% and +1.6% and +1.8%, respectively. Notably, our SNNs match or exceed ANNs' performance with the same architecture, and the energy efficiency is improved by 6.3${\times}$.

cs.NE↗

arXiv2025

DeformCL: Learning Deformable Centerline Representation for Vessel Extraction in 3D Medical Image

In the field of 3D medical imaging, accurately extracting and representing the blood vessels with curvilinear structures holds paramount importance for clinical diagnosis. Previous methods have commonly relied on discrete representation like mask, often resulting in local fractures or scattered fragments due to the inherent limitations of the per-pixel classification paradigm. In this work, we introduce DeformCL, a new continuous representation based on Deformable Centerlines, where centerline points act as nodes connected by edges that capture spatial relationships. Compared with previous representations, DeformCL offers three key advantages: natural connectivity, noise robustness, and interaction facility. We present a comprehensive training pipeline structured in a cascaded manner to fully exploit these favorable properties of DeformCL. Extensive experiments on four 3D vessel segmentation datasets demonstrate the effectiveness and superiority of our method. Furthermore, the visualization of curved planar reformation images validates the clinical significance of the proposed framework. We release the code in https://github.com/barry664/DeformCL

cs.CV↗

arXiv2025

Think Small, Act Big: Primitive Prompt Learning for Lifelong Robot Manipulation

Building a lifelong robot that can effectively leverage prior knowledge for continuous skill acquisition remains significantly challenging. Despite the success of experience replay and parameter-efficient methods in alleviating catastrophic forgetting problem, naively applying these methods causes a failure to leverage the shared primitives between skills. To tackle these issues, we propose Primitive Prompt Learning (PPL), to achieve lifelong robot manipulation via reusable and extensible primitives. Within our two stage learning scheme, we first learn a set of primitive prompts to represent shared primitives through multi-skills pre-training stage, where motion-aware prompts are learned to capture semantic and motion shared primitives across different skills. Secondly, when acquiring new skills in lifelong span, new prompts are appended and optimized with frozen pretrained prompts, boosting the learning via knowledge transfer from old skills to new ones. For evaluation, we construct a large-scale skill dataset and conduct extensive experiments in both simulation and real-world tasks, demonstrating PPL's superior performance over state-of-the-art methods.

cs.RO↗

arXiv2025

Entropic Diagram Characterization of Quantum Coherence: Degenerate Distillation and the Maximum Eigenvalue Uncertainty Bound

We develop a rigorous framework for quantifying quantum coherence in finite-dimensional systems by applying the Schur-Horn majorization theorem to relate eigenvalue distributions and diagonal entries of density matrices. Building on this foundation, we introduce a versatile suite of coherence measures, including the relative cross-entropy of coherence and its partial variants, that satisfy all resource theoretic axioms under incoherent operations. This unifying approach clarifies the geometric boundaries of physically realizable states in von Neumann-Tsallis entropy space and uncovers the phenomenon of degenerate coherence distillation where symmetry in the eigenvalue spectrum enables enhanced coherence extraction in higher-dimensional systems. In addition, we strengthen the entropy-based uncertainty relation by refining the Maassen-Uffink bound to account for the largest eigenvalues across distinct measurement bases. This refinement forges a deeper connection between entropy and uncertainty, which yields operationally meaningful constraints for quantum information tasks. Altogether, our findings illustrate the power of majorization in resource-theoretic analyses of quantum coherence, which offer valuable tools for both fundamental research and real-world applications in quantum information processing.

quant-ph↗

arXiv2025

An Investigation on Speaker Augmentation for End-to-End Speaker Extraction

Target confusion, defined as occasional switching to non-target speakers, poses a key challenge for end-to-end speaker extraction (E2E-SE) systems. We argue that this problem is largely caused by the lack of generalizability and discrimination of the speaker embeddings, and introduce a simple yet effective speaker augmentation strategy to tackle the problem. Specifically, we propose a time-domain resampling and rescaling pipeline that alters speaker traits while preserving other speech properties. This generates a variety of pseudo-speakers to help establish a generalizable speaker embedding space, while the speaker-trait-specific augmentation creates hard samples that force the model to focus on genuine speaker characteristics. Experiments on WSJ0-2Mix and LibriMix show that our method mitigates the target confusion and improves extraction performance. Moreover, it can be combined with metric learning, another effective approach to address target confusion, leading to further gains.

cs.SD↗