arXiv subjects

Jun Zhang

Publications and source records attributed to Jun Zhang.

At least 235 records · Page 13Linked to original sources

arXiv2026

DreamDojo: A Generalist Robot World Model from Large-Scale Human Videos

Being able to simulate the outcomes of actions in varied environments will revolutionize the development of generalist agents at scale. However, modeling these world dynamics, especially for dexterous robotics tasks, poses significant challenges due to limited data coverage and scarce action labels. As an endeavor towards this end, we introduce DreamDojo, a foundation world model that learns diverse interactions and dexterous controls from 44k hours of egocentric human videos. Our data mixture represents the largest video dataset to date for world model pretraining, spanning a wide range of daily scenarios with diverse objects and skills. To address the scarcity of action labels, we introduce continuous latent actions as unified proxy actions, enhancing interaction knowledge transfer from unlabeled videos. After post-training on small-scale target robot data, DreamDojo demonstrates a strong understanding of physics and precise action controllability. We also devise a distillation pipeline that accelerates DreamDojo to a real-time speed of 10.81 FPS and further improves context consistency. Our work enables several important applications based on generative world models, including live teleoperation, policy evaluation, and model-based planning. Systematic evaluation on multiple challenging out-of-distribution (OOD) benchmarks verifies the significance of our method for simulating open-world, contact-rich tasks, paving the way for general-purpose robot world models.

cs.RO↗

arXiv2026

Dynamics of string breaking and revival in a Rydberg atomic chain

String breaking is one of the most representative non-perturbative dynamics processes in confinement theory, typically associated with the creation of particle-antiparticle pairs. In this paper, we take a one-dimensional Rydberg atomic chain to theoretically study the dynamical of finite-length string state. Under different string tension conditions, we find that the string dynamics exhibits two clearly distinguishable evolution characteristics: one is that the string breaks and the system enters a superposition state space containing multiple meson state configurations; the other is localized string dynamics, in which the string undergoes local breaking but can then recombine and return to a state close to the initial structure, with the breaking and recombination processes recurring over a long time scale. Through the analysis of the evolution of different meson state configurations, we visually depict the redistribution of configuration weights during the string breaking process, and reveal the observable recovery characteristics of the string after breaking. Further analysis shows that the enhancement of quantum fluctuations can increase the weight of the double-meson state configurations in the system wave function without changing the dominant dynamical behavior. The above results present a rich picture of string breaking dynamics in a one-dimensional Rydberg atomic chain and provide insights for studying confinement physics and related gauge field theory phenomena on quantum simulation platforms.

cond-mat.quant-gas↗

arXiv2026

GAS: Enhancing Reward-Cost Balance of Generative Model-assisted Offline Safe RL

Offline Safe Reinforcement Learning (OSRL) aims to learn a policy to achieve high performance in sequential decision-making while satisfying constraints, using only pre-collected datasets. Recent works, inspired by the strong capabilities of Generative Models (GMs), reformulate decision-making in OSRL as a conditional generative process, where GMs generate desirable actions conditioned on predefined reward and cost values. However, GM-assisted methods face two major challenges in OSRL: (1) lacking the ability to "stitch" optimal transitions from suboptimal trajectories within the dataset, and (2) struggling to balance reward targets with cost targets, particularly when they are conflict. To address these issues, we propose Goal-Assisted Stitching (GAS), a novel algorithm designed to enhance stitching capabilities while effectively balancing reward maximization and constraint satisfaction. To enhance the stitching ability, GAS first augments and relabels the dataset at the transition level, enabling the construction of high-quality trajectories from suboptimal ones. GAS also introduces novel goal functions, which estimate the optimal achievable reward and cost goals from the dataset. These goal functions, trained using expectile regression on the relabeled and augmented dataset, allow GAS to accommodate a broader range of reward-cost return pairs and achieve a better tradeoff between reward maximization and constraint satisfaction compared to human-specified values. The estimated goals then guide policy training, ensuring robust performance under constrained settings. Furthermore, to improve training stability and efficiency, we reshape the dataset to achieve a more uniform reward-cost return distribution. Empirical results validate the effectiveness of GAS, demonstrating superior performance in balancing reward maximization and constraint satisfaction compared to existing methods.

cs.LG↗

arXiv2026

Large Language Model as Meta-Surrogate for Data-Driven Many-Task Optimization: A Proof-of-Principle Study

In many-task optimization scenarios, surrogate models are valuable for mitigating the computational burden of repeated fitness evaluations across tasks. This study proposes a novel meta-surrogate framework to assist many-task optimization, by leveraging the knowledge transfer strengths and emergent capabilities of large language models (LLMs). We formulate a unified framework for many-task fitness prediction, by defining a universal model with metadata to fit a group of problems. Fitness prediction is performed on metadata and decision variables, enabling efficient knowledge sharing across tasks and adaptability to new tasks. The LLM-based meta-surrogate treats fitness prediction as conditional probability estimation, employing a unified token sequence representation for task metadata, inputs, and outputs. This approach facilitates efficient inter-task knowledge sharing through shared token embeddings and captures complex task dependencies via multi-task model training. Experimental results demonstrate the model's emergent generalization ability, including zero-shot performance on problems with unseen dimensions. When integrated into evolutionary transfer optimization (ETO), our framework supports dual-level knowledge transfer -- at both the surrogate and individual levels -- enhancing optimization efficiency and robustness. This work establishes a novel foundation for applying LLMs in surrogate modeling, offering a versatile solution for many-task optimization.

cs.LG↗

arXiv2026

Common Envelope Evolution of Ultralight Boson Clouds

Ultralight bosons can be excited around spinning black holes via black hole superradiance. These boson clouds may play an important role in the orbital evolution of binary black holes. In this work, we investigate the formation and evolution of common envelopes of ultralight boson clouds in comparable mass-ratio black hole binaries. We describe the cloud evolution using gravitational molecular eigenstates and analyze the possible level transitions during orbital decay, as well as the impact on orbital dynamics. We find that the cloud can generally lead to eccentricity growth. In particular, the eccentricity may vary significantly during level transition, leaving an eccentricity of ${\cal O}(0.1)$ within the detection band of ground-based gravitational wave detectors.

gr-qc↗

arXiv2026

Focus-dLLM: Accelerating Long-Context Diffusion LLM Inference via Confidence-Guided Context Focusing

Diffusion Large Language Models (dLLMs) deliver strong long-context processing capability in a non-autoregressive decoding paradigm. However, the considerable computational cost of bidirectional full attention limits the inference efficiency. Although sparse attention is promising, existing methods remain ineffective. This stems from the need to estimate attention importance for tokens yet to be decoded, while the unmasked token positions are unknown during diffusion. In this paper, we present Focus-dLLM, a novel training-free attention sparsification framework tailored for accurate and efficient long-context dLLM inference. Based on the finding that token confidence strongly correlates across adjacent steps, we first design a past confidence-guided indicator to predict unmasked regions. Built upon this, we propose a sink-aware pruning strategy to accurately estimate and remove redundant attention computation, while preserving highly influential attention sinks. To further reduce overhead, this strategy reuses identified sink locations across layers, leveraging the observed cross-layer consistency. Experimental results show that our method offers more than $29\times$ lossless speedup under $32K$ context length. The code is publicly available at: https://github.com/Longxmas/Focus-dLLM

cs.CL↗

arXiv2026

Anisotropic time evolution of sound modes in Bjorken expanding holographic plasma

The speed of sound is a key parameter for characterizing equilibrium states. However, sound waves change their properties when propagating through rapidly evolving anisotropic media, such as the quark-gluon plasma created in heavy-ion collisions. This paper uses $\mathcal{N}=4$ Super-Yang-Mills theory to numerically study the time evolution of the speed and attenuation of sound modes along with the relaxation time in a plasma undergoing Bjorken expansion from various initial states in a quasi-static approximation. The longitudinal Bjorken expansion breaks the isotropy, resulting in two distinct sound speeds that range from just below the conformal value to the speed of light. An anisotropic hydrodynamic description is constructed and its applicability is discussed. Implications for the analysis of heavy ion data are considered.

nucl-th↗

arXiv2026

Embodied Intelligent Spectrum Management: A New Paradigm for Dynamic Spectrum Access

Wireless communication is evolving into an agent era, where numerous intelligent agents equipped with perception, reasoning, and interaction capabilities will operate in highly dynamic wireless environments. To complete diverse complex tasks, agent communication will play a critical role, which enables autonomous information exchange with external tools, services, and other agents. This trendy movement will dramatically increase spectrum demand and result in unprecedented challenges for spectrum management. However, current spectrum management paradigms, including static spectrum allocation and intelligent management, lack the flexibility and generalization to accommodate the dynamic and heterogeneous demands of agent communication. The recent advancements in embodied intelligence (EI) bring a promising solution, and this article will provide our vision of an emerging embodied intelligent spectrum management (EISM) paradigm. We start with an architecture for EISM, elaborating its key enabling technologies. Then, a prototype platform is presented to demonstrate the advantages of EISM. Finally, key challenges and open issues are outlined to facilitate future research in this emerging field.

cs.NI↗

arXiv2026

QVGen: Pushing the Limit of Quantized Video Generative Models

Video diffusion models (DMs) have enabled high-quality video synthesis. Yet, their substantial computational and memory demands pose serious challenges to real-world deployment, even on high-end GPUs. As a commonly adopted solution, quantization has proven notable success in reducing cost for image DMs, while its direct application to video DMs remains ineffective. In this paper, we present QVGen, a novel quantization-aware training (QAT) framework tailored for high-performance and inference-efficient video DMs under extremely low-bit quantization (e.g., 4-bit or below). We begin with a theoretical analysis demonstrating that reducing the gradient norm is essential to facilitate convergence for QAT. To this end, we introduce auxiliary modules ($Φ$) to mitigate large quantization errors, leading to significantly enhanced convergence. To eliminate the inference overhead of $Φ$, we propose a rank-decay strategy that progressively eliminates $Φ$. Specifically, we repeatedly employ singular value decomposition (SVD) and a proposed rank-based regularization $\mathbfγ$ to identify and decay low-contributing components. This strategy retains performance while zeroing out additional inference overhead. Extensive experiments across $4$ state-of-the-art (SOTA) video DMs, with parameter sizes ranging from $1.3\text{B}\sim14\text{B}$, show that QVGen is the first to reach full-precision comparable quality under 4-bit settings. Moreover, it significantly outperforms existing methods. For instance, our 3-bit CogVideoX-2B achieves improvements of $+25.28$ in Dynamic Degree and $+8.43$ in Scene Consistency on VBench. Code and models are available at https://github.com/ModelTC/QVGen.

cs.CV↗

arXiv2026

Modular Foundation Model Inference at the Edge: Network-Aware Microservice Optimization

Foundation models (FMs) unlock unprecedented multimodal and multitask intelligence, yet their cloud-centric deployment precludes real-time responsiveness and compromises user privacy. Meanwhile, monolithic execution at the edge remains infeasible under stringent resource limits and uncertain network dynamics. To bridge this gap, we propose a microservice-based FM inference framework that exploits the intrinsic functional asymmetry between heavyweight core services and agile light services. Our two-tier deployment strategy ensures robust Quality of Service (QoS) under resource contention. Specifically, core services are placed statically via a long-term network-aware integer program with sparsity constraints to form a fault-tolerant backbone. On the other hand, light services are orchestrated dynamically by a low-complexity online controller that integrates effective capacity theory with Lyapunov optimization, providing probabilistic latency guarantees under real-time workload fluctuations. Simulations demonstrate that our framework achieves over 84% average on-time task completion with moderate deployment costs and maintains strong robustness as the system load scales.

cs.DC↗

arXiv2026

Robust Grid-Forming Control Based on Virtual Flux Observer

This paper investigates the design and analysis of a novel grid-forming (GFM) control method for grid-connected converters (GCCs). The core novelty lies in a virtual flux observer-based synchronization and load angle control method. The terminal voltage of the converter is directly regulated to provide voltage-source behavior. The control parameters are designed for decoupling and pole placement. The proposed method exhibits strong robustness in stability and dynamical performance across varying and uncertain grid strengths. The robust control performance of the proposed method is first demonstrated by small-signal analysis, then validated by experiments on a 20 kVA power conversion system.

eess.SY↗

arXiv2026

Differential Privacy Image Generation with Reconstruction Loss and Noise Injection Using an Error Feedback SGD

Traditional data masking techniques such as anonymization cannot achieve the expected privacy protection while ensuring data utility for privacy-preserving machine learning. Synthetic data plays an increasingly important role as it generates a large number of training samples and prevents information leakage in real data. The existing methods suffer from the repeating trade-off processes between privacy and utility. We propose a novel framework for differential privacy generation, which employs an Error Feedback Stochastic Gradient Descent(EFSGD) method and introduces a reconstruction loss and noise injection mechanism into the training process. We generate images with higher quality and usability under the same privacy budget as the related work. Extensive experiments demonstrate the effectiveness and generalization of our proposed framework for both grayscale and RGB images. We achieve state-of-the-art results over almost all metrics on three benchmarks: MNIST, Fashion-MNIST, and CelebA.

cs.CV↗

arXiv2026

KNN-SSD: Enabling Dynamic Self-Speculative Decoding via Nearest Neighbor Layer Set Optimization

Speculative Decoding (SD) has emerged as a widely used paradigm to accelerate the inference of large language models (LLMs) without compromising generation quality. It works by efficiently drafting multiple tokens using a compact model and then verifying them in parallel using the target LLM. Notably, Self-Speculative Decoding proposes skipping certain layers to construct the draft model, which eliminates the need for additional parameters or training. Despite its strengths, we observe in this work that drafting with layer skipping exhibits significant sensitivity to domain shifts, leading to a substantial drop in acceleration performance. To enhance the domain generalizability of this paradigm, we introduce KNN-SSD, an algorithm that leverages K-Nearest Neighbor (KNN) search to match different skipped layers with various domain inputs. We evaluated our algorithm in various models and multiple tasks, observing that its application leads to 1.3x-1.6x speedup in LLM inference.

cs.CL↗

arXiv2026

Large AI Models for Wireless Physical Layer

Large artificial intelligence models (LAMs) are transforming wireless physical layer technologies through their robust generalization, multitask processing, and multimodal capabilities. This article reviews recent advancements in applying LAMs to physical layer communications, addressing obstacles of conventional AI-based approaches. LAM-based solutions are classified into two strategies: leveraging pre-trained LAMs and developing native LAMs designed specifically for physical layer tasks. The motivations and key frameworks of these approaches are comprehensively examined through multiple use cases. Both strategies significantly improve performance and adaptability across diverse wireless scenarios. Future research directions, including efficient architectures, interpretability, standardized datasets, and collaboration between large and small models, are proposed to advance LAM-based physical layer solutions for next-generation communication systems.

cs.IT↗

arXiv2026

Joint analysis of small-scale galaxy clustering and galaxy--galaxy lensing from BOSS galaxies

We present a joint analysis of galaxy clustering and galaxy--galaxy lensing measurements from BOSS galaxies using a simulation-based emulation method combined with a halo occupation distribution model. Our emulators are constructed with the Aemulus $ν$ simulations, a suite of $wν$CDM $N$-body simulations with massive neutrinos as independent particle species. We combine small-scale analysis of clustering from $0.1h^{-1}$Mpc to $60.2~h^{-1}$Mpc and lensing from $1.7h^{-1}$Mpc to $60.2~h^{-1}$Mpc to perform cosmological constraints. We split the BOSS galaxies into three redshift bins to measure their clustering and employ galaxies from Dark Energy Camera Legacy Survey and Hyper Suprime-Cam as source galaxies to measure lensing separately. We find that the addition of lensing significantly improves the constraining power on $S_{8}=σ_8(Ω_m/0.3)^{0.5}$, with a weak improvement for $fσ_{8}$. Our results of $fσ_{8}$ indicate tensions of around $1\sim4σ$ below the results of CMB observations of Planck. For $S_{8}$, our results are also lower than Planck, and the tension can be mitigated when considering possible systematics in lensing measurement. As a byproduct, our analysis prefers a non-zero neutrino mass but without strong significance, with the constraining power dominated by the clustering. Given the accuracy and precision of our model and the observational data, it is anticipated that larger and higher-quality spectroscopic datasets will improve the constraints on this fundamental property in the near future.

astro-ph.CO↗

arXiv2026

STEP3-VL-10B Technical Report

We present STEP3-VL-10B, a lightweight open-source foundation model designed to redefine the trade-off between compact efficiency and frontier-level multimodal intelligence. STEP3-VL-10B is realized through two strategic shifts: first, a unified, fully unfrozen pre-training strategy on 1.2T multimodal tokens that integrates a language-aligned Perception Encoder with a Qwen3-8B decoder to establish intrinsic vision-language synergy; and second, a scaled post-training pipeline featuring over 1k iterations of reinforcement learning. Crucially, we implement Parallel Coordinated Reasoning (PaCoRe) to scale test-time compute, allocating resources to scalable perceptual reasoning that explores and synthesizes diverse visual hypotheses. Consequently, despite its compact 10B footprint, STEP3-VL-10B rivals or surpasses models 10$\times$-20$\times$ larger (e.g., GLM-4.6V-106B, Qwen3-VL-235B) and top-tier proprietary flagships like Gemini 2.5 Pro and Seed-1.5-VL. Delivering best-in-class performance, it records 92.2% on MMBench and 80.11% on MMMU, while excelling in complex reasoning with 94.43% on AIME2025 and 75.95% on MathVision. We release the full model suite to provide the community with a powerful, efficient, and reproducible baseline.

cs.CV↗

arXiv2026

Selecting and Testing Asset Pricing Models: A Stepwise Approach

The asset pricing literature emphasizes factor models that minimize pricing errors but overlooks unselected candidate factors that could enhance the performance of test assets. This paper proposes a framework for factor model selection and testing by (i) selecting the optimal model that spans the joint efficient frontier of test assets and all candidate factors, and (ii) testing pricing performance on both test assets and unselected candidate factors. Our framework updates a baseline model (e.g., CAPM) sequentially by adding or removing factors based on asset pricing tests. Ensuring model selection consistency, our framework utilizes the asset pricing duality: minimizing cross-sectionally unexplained pricing errors aligns with maximizing the Sharpe ratio of the selected factor model. Empirical evidence shows that workhorse factor models fail asset pricing tests, whereas our proposed 8-factor model is not rejected and exhibits robust out-of-sample performance.

econ.EM↗

arXiv2026

Distribution-Aligned Sequence Distillation for Superior Long-CoT Reasoning

In this report, we introduce DASD-4B-Thinking, a lightweight yet highly capable, fully open-source reasoning model. It achieves SOTA performance among open-source models of comparable scale across challenging benchmarks in mathematics, scientific reasoning, and code generation -- even outperforming several larger models. We begin by critically reexamining a widely adopted distillation paradigm in the community: SFT on teacher-generated responses, also known as sequence-level distillation. Although a series of recent works following this scheme have demonstrated remarkable efficiency and strong empirical performance, they are primarily grounded in the SFT perspective. Consequently, these approaches focus predominantly on designing heuristic rules for SFT data filtering, while largely overlooking the core principle of distillation itself -- enabling the student model to learn the teacher's full output distribution so as to inherit its generalization capability. Specifically, we identify three critical limitations in current practice: i) Inadequate representation of the teacher's sequence-level distribution; ii) Misalignment between the teacher's output distribution and the student's learning capacity; and iii) Exposure bias arising from teacher-forced training versus autoregressive inference. In summary, these shortcomings reflect a systemic absence of explicit teacher-student interaction throughout the distillation process, leaving the essence of distillation underexploited. To address these issues, we propose several methodological innovations that collectively form an enhanced sequence-level distillation training pipeline. Remarkably, DASD-4B-Thinking obtains competitive results using only 448K training samples -- an order of magnitude fewer than those employed by most existing open-source efforts. To support community research, we publicly release our models and the training dataset.

cs.LG↗