arXiv subjects

Peng Chen

Publications and source records attributed to Peng Chen.

At least 109 records · Page 6Linked to original sources

arXiv2026

CombatVLA: An Efficient Vision-Language-Action Model for Combat Tasks in 3D Action Role-Playing Games

Recent advances in Vision-Language-Action models (VLAs) have expanded the capabilities of embodied intelligence. However, significant challenges remain in real-time decision-making in complex 3D environments, which demand second-level responses, high-resolution perception, and tactical reasoning under dynamic conditions. To advance the field, we introduce CombatVLA, an efficient VLA model optimized for combat tasks in 3D action role-playing games(ARPGs). Specifically, our CombatVLA is a 3B model trained on video-action pairs collected by an action tracker, where the data is formatted as action-of-thought (AoT) sequences. Thereafter, CombatVLA seamlessly integrates into an action execution framework, allowing efficient inference through our truncated AoT strategy. Experimental results demonstrate that CombatVLA not only outperforms all existing models on the combat understanding benchmark but also achieves a 50-fold acceleration in game combat. Moreover, it has a higher task success rate than human players. We will open-source all resources, including the action tracker, dataset, benchmark, model weights, training code, and the implementation of the framework at https://combatvla.github.io/.

cs.CV↗

arXiv2026

Rotate Your Character: Revisiting Video Diffusion Models for High-Quality 3D Character Generation

Generating high-quality 3D characters from single images remains a significant challenge in digital content creation, particularly due to complex body poses and self-occlusion. In this paper, we present RCM (Rotate your Character Model), an advanced image-to-video diffusion framework tailored for high-quality novel view synthesis (NVS) and 3D character generation. Compared to existing diffusion-based approaches, RCM offers several key advantages: (1) transferring characters with any complex poses into a canonical pose, enabling consistent novel view synthesis across the entire viewing orbit, (2) high-resolution orbital video generation at 1024x1024 resolution, (3) controllable observation positions given different initial camera poses, and (4) multi-view conditioning supporting up to 4 input images, accommodating diverse user scenarios. Extensive experiments demonstrate that RCM outperforms state-of-the-art methods in both novel view synthesis and 3D generation quality.

cs.CV↗

arXiv2026

Sequential Bayesian Optimal Experimental Design in Infinite Dimensions via Policy Gradient Reinforcement Learning

Sequential Bayesian optimal experimental design (SBOED) for PDE-governed inverse problems is computationally challenging, especially for infinite-dimensional random field parameters. High-fidelity approaches require repeated forward and adjoint PDE solves inside nested Bayesian inversion and design loops. We formulate SBOED as a finite-horizon Markov decision process and learn an amortized design policy via policy-gradient reinforcement learning (PGRL), enabling online design selection from the experiment history without repeatedly solving an SBOED optimization problem. To make policy training and reward evaluation scalable, we combine dual dimension reduction -- active subspace projection for the parameter and principal component analysis for the state -- with an adjusted derivative-informed latent attention neural operator (LANO) surrogate that predicts both the parameter-to-solution map and its Jacobian. We use a Laplace-based D-optimality reward while noting that, in general, other expected-information-gain utilities such as KL divergence can also be used within the same framework. We further introduce an eigenvalue-based evaluation strategy that uses prior samples as proxies for maximum a posteriori (MAP) points, avoiding repeated MAP solves while retaining accurate information-gain estimates. Numerical experiments on sequential multi-sensor placement for contaminant source tracking demonstrate approximately $100\times$ speedup over high-fidelity finite element methods, improved performance over random sensor placements, and physically interpretable policies that discover an ``upstream'' tracking strategy.

math.OC↗

arXiv2026

AT$^2$PO: Agentic Turn-based Policy Optimization via Tree Search

LLM agents have emerged as powerful systems for tackling multi-turn tasks by interleaving internal reasoning and external tool interactions. Agentic Reinforcement Learning has recently drawn significant research attention as a critical post-training paradigm to further refine these capabilities. In this paper, we present AT$^2$PO (Agentic Turn-based Policy Optimization via Tree Search), a unified framework for multi-turn agentic RL that addresses three core challenges: limited exploration diversity, sparse credit assignment, and misaligned policy optimization. AT$^2$PO introduces a turn-level tree structure that jointly enables Entropy-Guided Tree Expansion for strategic exploration and Turn-wise Credit Assignment for fine-grained reward propagation from sparse outcomes. Complementing this, we propose Agentic Turn-based Policy Optimization, a turn-level learning objective that aligns policy updates with the natural decision granularity of agentic interactions. ATPO is orthogonal to tree search and can be readily integrated into any multi-turn RL pipeline. Experiments across seven benchmarks demonstrate consistent improvements over the state-of-the-art baseline by up to 1.84 percentage points in average, with ablation studies validating the effectiveness of each component. Our code is available at https://github.com/zzfoutofspace/ATPO.

cs.AI↗

arXiv2026

KOS-TL (Knowledge Operation System Type Logic)

This paper introduces KOS-TL (Knowledge Operation System Type Logic), a novel constructive framework designed to provide a rigorous logical foundation for autonomous and executable knowledge systems. Traditional knowledge representation models often suffer from a gap between static symbolic logic and dynamic system execution. To bridge this divide, KOS-TL leverages Dependent Type Theory to unify data, logic, and proof into a singular computational substrate.The architecture of KOS-TL is organized into three hierarchical layers: the Core Layer, which defines the static type universe and constructive primitives; the Kernel Layer, which governs state evolution through an event-driven mechanism characterized by the triple $\langle Σ, \textsf{Ev}, Δ\rangle$; and the Runtime Layer, responsible for the bidirectional refinement of physical signals into logical evidence. We formally define the operational semantics of the system and prove key meta-theoretical properties, including Progress and Evolutionary Consistency, ensuring that the system remains logically self-consistent and free from stuck states during continuous state transitions.By integrating Davidsonian event semantics with Martin-Löf type theory, KOS-TL enables the construction of "proof-carrying knowledge," where every state change in the knowledge base is accompanied by a formal witness of its validity. We demonstrate the practical utility of this logic through application examples in industrial traceability and cross-border financial compliance. Our results suggest that KOS-TL provides a robust, formally verifiable basis for the next generation of intelligent, autonomous operating systems.

cs.CL↗

arXiv2025

BIOME-Bench: A Benchmark for Biomolecular Interaction Inference and Multi-Omics Pathway Mechanism Elucidation from Scientific Literature

Multi-omics studies often rely on pathway enrichment to interpret heterogeneous molecular changes, but pathway enrichment (PE)-based workflows inherit structural limitations of pathway resources, including curation lag, functional redundancy, and limited sensitivity to molecular states and interventions. Although recent work has explored using large language models (LLMs) to improve PE-based interpretation, the lack of a standardized benchmark for end-to-end multi-omics pathway mechanism elucidation has largely confined evaluation to small, manually curated datasets or ad hoc case studies, hindering reproducible progress. To address this issue, we introduce BIOME-Bench, constructed via a rigorous four-stage workflow, to evaluate two core capabilities of LLMs in multi-omics analysis: Biomolecular Interaction Inference and end-to-end Multi-Omics Pathway Mechanism Elucidation. We develop evaluation protocols for both tasks and conduct comprehensive experiments across multiple strong contemporary models. Experimental results demonstrate that existing models still exhibit substantial deficiencies in multi-omics analysis, struggling to reliably distinguish fine-grained biomolecular relation types and to generate faithful, robust pathway-level mechanistic explanations.

cs.CL↗

arXiv2025

Variationally correct operator learning: Reduced basis neural operator with a posteriori error estimation

Minimizing PDE-residual losses is a common strategy to promote physical consistency in neural operators. However, standard formulations often lack variational correctness, meaning that small residuals do not guarantee small solution errors due to the use of non-compliant norms or ad hoc penalty terms for boundary conditions. This work develops a variationally correct operator learning framework by constructing first-order system least-squares (FOSLS) objectives whose values are provably equivalent to the solution error in PDE-induced norms. We demonstrate this framework on stationary diffusion and linear elasticity, incorporating mixed Dirichlet-Neumann boundary conditions via variational lifts to preserve norm equivalence without inconsistent penalties. To ensure the function space conformity required by the FOSLS loss, we propose a Reduced Basis Neural Operator (RBNO). The RBNO predicts coefficients for a pre-computed, conforming reduced basis, thereby ensuring variational stability by design while enabling efficient training. We provide a rigorous convergence analysis that bounds the total error by the sum of finite element discretization bias, reduced basis truncation error, neural network approximation error, and statistical estimation errors arising from finite sampling and optimization. Numerical benchmarks validate these theoretical bounds and demonstrate that the proposed approach achieves superior accuracy in PDE-compliant norms compared to standard baselines, while the residual loss serves as a reliable, computable a posteriori error estimator.

math.NA↗

arXiv2025

PeerSync: Accelerating Containerized Service Delivery at the Network Edge

Efficient container image distribution is crucial for enabling machine learning inference at the network edge, where resource limitations and dynamic network conditions create significant challenges. In this paper, we present PeerSync, a decentralized P2P-based system designed to optimize image distribution in edge environments. PeerSync employs a popularity- and network-aware download engine that dynamically adapts to content popularity and real-time network conditions. PeerSync further integrates automated tracker election for rapid peer discovery and dynamic cache management for efficient storage utilization. We implement PeerSync with 8000+ lines of Rust code and test its performance extensively on both large-scale Docker-based emulations and physical edge devices. Experimental results show that PeerSync delivers a remarkable speed increase of 2.72$\times$, 1.79$\times$, and 1.28$\times$ compared to the Baseline solution, Dragonfly, and Kraken, respectively, while significantly reducing cross-network traffic by 90.72% under congested and varying network conditions.

cs.NI↗

arXiv2025

In-plane optically tunable magnetic states in 2D materials via tailored femtosecond laser driving

It is well established that light can control magnetism in matter, e.g. via the inverse Faraday effect or ultrafast demagnetization. However, such control is typically limited to magnetization transverse to light's polarization plane, or out-of-plane magnetism in 2D materials, while in-plane magnetic moments have remained largely unexplored. This is due to the difficulty of generating electronic orbital angular momentum components within light's polarization plane. Here we overcome this limitation, demonstrating complete three-dimensional, all-optical control of magnetism in 2D materials. Using first-principles simulations, we show that a tailored, two-color laser field can induce and steer magnetic moments in any direction with the relative angle between the laser polarizations playing a key parameter in coherent control. We analyze the physical mechanism of this process and show that it arises from a simultaneous breaking of time-reversal and spatial-inversion symmetries in the two-color laser. In-plane orbital moments are introduced via non-zero out-of-plane longitudinal photogalvanic currents enabled by broken inversion and mirror symmetries, while time-reversal symmetry breaking enables build-up of spin-rotation processes through spin-orbit coupling, translating the orbital moments to transient magnetism. Our findings demonstrate a full 3D coherent control scheme for transient magnetic states on femtosecond timescales driven by tailored lasers, and can be used to develop novel spectroscopies for magnetism, all-optical magnetic switching for ultrafast spintronics, and novel information storage capabilities.

physics.optics↗

arXiv2025

PDPO: Parametric Density Path Optimization

We introduce Parametric Density Path Optimization (PDPO), a novel method for computing action-minimizing paths between probability densities. The core idea is to represent the target probability path as the pushforward of a reference density through a parametric map, transforming the original infinite-dimensional optimization over densities to a finite-dimensional one over the parameters of the map. We derive a static formulation of the dynamic problem of action minimization and propose cubic spline interpolation of the path in parameter space to solve the static problem. Theoretically, we establish an error bound of the action under proper assumptions on the regularity of the parameter path. Empirically, we find that using 3-5 control points of the spline interpolation suffices to accurately resolve both multimodal and high-dimensional problems. We demonstrate that PDPO can flexibly accommodate a wide range of potential terms, including those modeling obstacles, mean-field interactions, stochastic control, and higher-order dynamics. Our method outperforms existing state-of-the-art approaches in benchmark tasks, demonstrating superior computational efficiency and solution quality. The source code will be publically available after the revision process.

math.OC↗

arXiv2025

Towards Non-Stationary Time Series Forecasting with Temporal Stabilization and Frequency Differencing

Time series forecasting is critical for decision-making across dynamic domains such as energy, finance, transportation, and cloud computing. However, real-world time series often exhibit non-stationarity, including temporal distribution shifts and spectral variability, which pose significant challenges for long-term time series forecasting. In this paper, we propose DTAF, a dual-branch framework that addresses non-stationarity in both the temporal and frequency domains. For the temporal domain, the Temporal Stabilizing Fusion (TFS) module employs a non-stationary mix of experts (MOE) filter to disentangle and suppress temporal non-stationary patterns while preserving long-term dependencies. For the frequency domain, the Frequency Wave Modeling (FWM) module applies frequency differencing to dynamically highlight components with significant spectral shifts. By fusing the complementary outputs of TFS and FWM, DTAF generates robust forecasts that adapt to both temporal and frequency domain non-stationarity. Extensive experiments on real-world benchmarks demonstrate that DTAF outperforms state-of-the-art baselines, yielding significant improvements in forecasting accuracy under non-stationary conditions. All codes are available at https://github.com/decisionintelligence/DTAF.

cs.LG↗

arXiv2025

HunyuanVideo 1.5 Technical Report

We present HunyuanVideo 1.5, a lightweight yet powerful open-source video generation model that achieves state-of-the-art visual quality and motion coherence with only 8.3 billion parameters, enabling efficient inference on consumer-grade GPUs. This achievement is built upon several key components, including meticulous data curation, an advanced DiT architecture featuring selective and sliding tile attention (SSTA), enhanced bilingual understanding through glyph-aware text encoding, progressive pre-training and post-training, and an efficient video super-resolution network. Leveraging these designs, we developed a unified framework capable of high-quality text-to-video and image-to-video generation across multiple durations and resolutions. Extensive experiments demonstrate that this compact and proficient model establishes a new state-of-the-art among open-source video generation models. By releasing the code and model weights, we provide the community with a high-performance foundation that lowers the barrier to video creation and research, making advanced video generation accessible to a broader audience. All open-source assets are publicly available at https://github.com/Tencent-Hunyuan/HunyuanVideo-1.5.

cs.CV↗

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

A segment anchoring-based balancing algorithm for agricultural multi-robot task allocation with energy constraints

Multi-robot systems have emerged as a key technology for addressing the efficiency and cost challenges in labor-intensive industries. In the representative scenario of smart farming, planning efficient harvesting schedules for a fleet of electric robots presents a highly challenging frontier problem. The complexity arises not only from the need to find Pareto-optimal solutions for the conflicting objectives of makespan and transportation cost, but also from the necessity to simultaneously manage payload constraints and finite battery capacity. When robot loads are dynamically updated during planned multi-trip operations, a mandatory recharge triggered by energy constraints introduces an unscheduled load reset. This interaction creates a complex cascading effect that disrupts the entire schedule and renders traditional optimization methods ineffective. To address this challenge, this paper proposes the segment anchoring-based balancing algorithm (SABA). The core of SABA lies in the organic combination of two synergistic mechanisms: the sequential anchoring and balancing mechanism, which leverages charging decisions as `anchors' to systematically reconstruct disrupted routes, while the proportional splitting-based rebalancing mechanism is responsible for the fine-grained balancing and tuning of the final solutions' makespans. Extensive comparative experiments, conducted on a real-world case study and a suite of benchmark instances, demonstrate that SABA comprehensively outperforms 6 state-of-the-art algorithms in terms of both solution convergence and diversity. This research provides a novel theoretical perspective and an effective solution for the multi-robot task allocation problem under energy constraints.

cs.MA↗

arXiv2025

An adaptive experience-based discrete genetic algorithm for multi-trip picking robot task scheduling in smart orchards

The continuous innovation of smart robotic technologies is driving the development of smart orchards, significantly enhancing the potential for automated harvesting systems. While multi-robot systems offer promising solutions to address labor shortages and rising costs, the efficient scheduling of these systems presents complex optimization challenges. This research investigates the multi-trip picking robot task scheduling (MTPRTS) problem. The problem is characterized by its provision for robot redeployment while maintaining strict adherence to makespan constraints, and encompasses the interdependencies among robot weight, robot load, and energy consumption, thus introducing substantial computational challenges that demand sophisticated optimization algorithms.To effectively tackle this complexity, metaheuristic approaches, which often utilize local search mechanisms, are widely employed. Despite the critical role of local search in vehicle routing problems, most existing algorithms are hampered by redundant local operations, leading to slower search processes and higher risks of local optima, particularly in large-scale scenarios. To overcome these limitations, we propose an adaptive experience-based discrete genetic algorithm (AEDGA) that introduces three key innovations: (1) integrated load-distance balancing initialization method, (2) a clustering-based local search mechanism, and (3) an experience-based adaptive selection strategy. To ensure solution feasibility under makespan constraints, we develop a solution repair strategy implemented through three distinct frameworks. Comprehensive experiments on 18 proposed test instances and 24 existing test problems demonstrate that AEDGA significantly outperforms eight state-of-the-art algorithms.

cs.RO↗

arXiv2025

Towards a deeper fundamental understanding of (Al,Sc)N ferroelectric nitrides

Density Functional Theory (DFT) calculations, within the virtual crystal alloy approximation, are performed, along with the development of a Landau-type model employing a symmetry-allowed analytical expression of the internal energy and having parameters being determined from first principles, to investigate properties and energetics of Al1-xScxN ferroelectric nitrides in their hexagonal forms. These DFT computations and this model predict the existence of two different types of minima, namely the 4-fold-coordinated wurtzite (WZ) polar structure and a 5-times paraelectric hexagonal phase (to be denoted as H5), for any Sc composition up to 40%. The H5 minimum progressively becomes the lowest energy state within hexagonal symmetry as the Sc concentration increases from 0 to 40%. Furthermore, the model points out to several key findings. Examples include the crucial role of the coupling between polarization and strains to create the WZ minimum, in addition to polar and elastic energies, and that the origin of the H5 state overcoming the WZ phase as the global minimum within hexagonal symmetry when increasing the Sc composition mostly lies in the compositional dependency of only two parameters, one linked to the polarization and another one being purely elastic in nature. Other examples are that forcing Al1-xScxN systems to have no or a weak change in lattice parameters when heating them allows to reproduce well their finite-temperature polar properties, and that a value of the axial ratio close to that of the ideal WZ structure does imply a large polarization at low temperatures but not necessarily at high temperatures because of the ordered-disordered character of the temperature-induced formation of the WZ state. Such findings should allow for a better fundamental understanding of (Al,Sc)N ferroelectric nitrides, which may be used to design efficient devices operating at low voltages.

cond-mat.mtrl-sci↗

arXiv2025

Robustifying Learning-Augmented Caching Efficiently without Compromising 1-Consistency

The online caching problem aims to minimize cache misses when serving a sequence of requests under a limited cache size. While naive learning-augmented caching algorithms achieve ideal $1$-consistency, they lack robustness guarantees. Existing robustification methods either sacrifice $1$-consistency or introduce excessive computational overhead. In this paper, we introduce Guard, a lightweight robustification framework that enhances the robustness of a broad class of learning-augmented caching algorithms to $2H_{k-1} + 2$, while preserving their $1$-consistency. Guard achieves the current best-known trade-off between consistency and robustness, with only O(1) additional per-request overhead, thereby maintaining the original time complexity of the base algorithm. Extensive experiments across multiple real-world datasets and prediction models validate the effectiveness of Guard in practice.

cs.DS↗

arXiv2025

Three-dimensional flat bands and possible interlayer triplet pairing superconductivity in the alternating twisted NbSe$_2$ moiré bulk

Moiré superlattices hosting flat bands and correlated states have emerged as a focal topic in condensed matter research. Through first-principles calculations, we investigate three-dimensional flat bands in alternating twisted NbSe$_2$ moiré bulk structures. These structures exhibit enhanced interlayer interactions compared to twisted bilayer configurations. Our results demonstrate that moiré bulks undergo spontaneous large-scale structural relaxation, resulting in the formation of remarkably flat energy bands at twist angles $\leq$ 7.31°. The $k_z$-dependent dispersion of flat bands across different moiré bulks reveals their intrinsic three-dimensional character. The presence of out-of-plane mirror symmetry in these moiré bulk structures suggests possible interlayer triplet superconducting pairing mechanisms that differ from those in twisted bilayer systems. Our work paves the way for exploring potential three-dimensional flat bands in other moiré bulk systems.

cond-mat.supr-con↗