2026-07论文中文摘要
用于检索增强生成系统的交叉注意力校准去重
Cross-Attention Calibrated Deduplication for Retrieval-Augmented Generation System
arXiv 2607.24332 · 2026-07-28DynaCalKV：通过头部分组和自适应秩分配实现键值缓存压缩
DynaCalKV: Key-Value Cache Compression via Head Grouping and Adaptive Rank Allocation
arXiv 2607.24331 · 2026-07-28迈向中高频段系统中无别名信道外推：一种空间 - 频率 - 时间张量学习方法
Toward Alias-Free Channel Extrapolation in Upper Mid-Band Systems: A Spatial-Frequency-Temporal Tensor Learning Approach
arXiv 2607.24330 · 2026-07-28紫外边界条件与希格斯质量
Ultraviolet boundary condition and the Higgs mass
arXiv 2607.24329 · 2026-07-28对齐异构DFT数据集：一种用于交叉泛函形成能的图神经网络方法
Aligning Heterogeneous DFT Datasets: A Graph Neural Network Approach to Cross-Functional Formation Energies
arXiv 2607.24327 · 2026-07-28由输运噪声驱动的随机相分离
Stochastic phase separation driven by transport noise
arXiv 2607.24326 · 2026-07-28用于双曲守恒律的基于五阶有限体积导数的具有统一模板的Hermite WENO格式
Fifth-order finite volume derivative-based Hermite WENO scheme with unified stencils for hyperbolic conservation laws
arXiv 2607.24325 · 2026-07-28高维线性模型的扩散自举法
Diffusion Bootstrap for High-Dimensional Linear Models
arXiv 2607.24324 · 2026-07-28重新审视Vocos：时频神经声码器中的相位问题
Revisiting Vocos: That Phasiness Business in Time-Frequency Neural Vocoding
arXiv 2607.24323 · 2026-07-28随机流形的欧拉特征
Euler Characteristics of Random Manifolds
arXiv 2607.24322 · 2026-07-28用于与现有高能物理代码库集成的Julia包的静态编译：以JetReconstruction.jl为例
Static compilation of Julia packages for integration with existing HEP codebases: a case study with JetReconstruction.jl
arXiv 2607.24321 · 2026-07-28用于RoboRacer平台自主赛车中泛化的持续强化学习
Continual-RL for Generalization in Autonomous Racing on the RoboRacer Platform
arXiv 2607.24320 · 2026-07-28具有压缩的全组态相互作用坐标下降法的收敛性与误差分析
The Convergence and Error Analysis of Coordinate Descent Methods with Compression for Full Configuration Interaction
arXiv 2607.24319 · 2026-07-28快速生成计量相关的福克态混合态
Fast Generation of Metrologically Relevant Fock State Mixtures
arXiv 2607.24318 · 2026-07-28关于约束刚体时间步长方案中运动表示和构型更新的一个注记
A note on the motion representation and configuration update in time stepping schemes for the constrained rigid body
arXiv 2607.24317 · 2026-07-28通过Bruhat-Tits树扩展的广义Vladimirov导数的能量关系
Energy relations for the generalised Vladimirov derivative via Bruhat-Tits tree extensions
arXiv 2607.24316 · 2026-07-28研究B→PPℓ⁺ℓ⁻和B→PVℓ⁺ℓ⁻衰变中的张量共振贡献
Studying the tensor resonance contributions in $B \to PP\ell^+\ell^-$ and $B \to PV\ell^+\ell^-$ decays
arXiv 2607.24315 · 2026-07-28MEGA-CL：通过图外部注意力和对比学习实现可泛化ADMET预测的分子基础模型
MEGA-CL: A Molecular Foundation Model for Generalizable ADMET Prediction through Graph External Attention and Contrastive Learning
arXiv 2607.24314 · 2026-07-28通过局部多智能体RAG实现能量受限的分层水下监测
Energy Constrained Hierarchical Underwater Monitoring via Local Multi-Agent RAG
arXiv 2607.24313 · 2026-07-28通过二次特征的带有克洛斯特曼和的双线性形式
Bilinear forms with Kloosterman sums via quadratic characters
arXiv 2607.24311 · 2026-07-28记忆算子集合表明在模拟大西洋经向翻转环流转变中接近临界状态
Memory operator ensembles indicate proximity to criticality in simulated AMOC transitions
arXiv 2607.24310 · 2026-07-28关于Werner态的双拷贝可蒸馏性及一个新的部分跟踪不等式
On the two-copy distillability of Werner states and a new partial trace inequality
arXiv 2607.24309 · 2026-07-28重新思考块扩散语言模型的生成顺序
Rethinking the Generation Order of Block Diffusion Language Models
arXiv 2607.24306 · 2026-07-28用于高维随机偏微分方程中基于可扩展分散集合卡尔曼滤波器的参数识别的域分解神经代理
Domain-Decomposition Neural Surrogates for Scalable Decentralized Ensemble Kalman Filter Based Parameter Identification in High-Dimensional Stochastic PDEs
arXiv 2607.24305 · 2026-07-28教师最了解：网络朗之万动力学人工智能谄媚中的自发对称性破缺和临界点
Teacher Knows It Best: Spontaneous Symmetry Breaking and Tipping Points in Networked Langevin Dynamics AI Sycophancy
arXiv 2607.24304 · 2026-07-28具有硬件损伤的高空平台站辅助太赫兹卫星通信系统
High-Altitude Platform Station-Aided Terahertz Satellite Communication Systems with Hardware Impairments
arXiv 2607.24303 · 2026-07-28大场景下有遮挡的多视角多人人体网格恢复
Multiview Multi-Person Human Mesh Recovery Under Large Scenes with Occlusions
arXiv 2607.24302 · 2026-07-28离散时间相位Petri盒演算dtphPBC
Discrete time phased Petri box calculus dtphPBC
arXiv 2607.24301 · 2026-07-28在启发式自我改进智能体中自我编写的验证不可靠
Self-Authored Verification Is Unreliable in Heuristic Self-Improving Agents
arXiv 2607.24300 · 2026-07-28UMI3D：通过同步聚焦交叉注意力路由在无约束多图像输入上进行稳健的3D生成
UMI3D: Robust 3D Generation on Unconstrained Multi-Image Inputs via Simultaneous Focus Cross-Attention Routing
arXiv 2607.24298 · 2026-07-28关于队列数为1的外平面图的识别
On the Recognition of Outerplanar Graphs with Queue Number 1
arXiv 2607.24297 · 2026-07-28PAC-DP：PAC贝叶斯扩散策略学习
PAC-DP: PAC-Bayesian Diffusion Policy Learning
arXiv 2607.24296 · 2026-07-28秩度量中代码局部性的新视角
New perspectives for code locality in the rank metric
arXiv 2607.24295 · 2026-07-28测量强度调制器Vpi的方法之间的比较
Comparison between methods to measureintensity modulator Vpi
arXiv 2607.24294 · 2026-07-28具有黎曼预测器的局部弗雷歇回归
Local Fréchet Regression with Riemannian Predictors
arXiv 2607.24293 · 2026-07-28通过超网络学习自适应多任务制导、导航与控制
Learning Adaptive Multi-Task Guidance, Navigation, and Control via Hypernetworks
arXiv 2607.24292 · 2026-07-28RPG-VST：用于盲 RAW 去噪的鲁棒泊松-高斯方差稳定化
RPG-VST: Robust Poisson-Gaussian Variance Stabilization for Blind RAW Denoising
arXiv 2607.24291 · 2026-07-28来自普朗克巡天的（亚）毫米波波段星系的局部光度函数
Local luminosity functions of galaxies at (sub)millimeter wavelengths from Planck surveys
arXiv 2607.24289 · 2026-07-28基于超像素的二次无约束二进制优化用于可扩展量子增强医学图像分割
Superpixel-Based QUBO for Scalable Quantum-Enhanced Medical Image Segmentation
arXiv 2607.24288 · 2026-07-28语义最小能量原理：一种关于智能的假设
The Semantic Least-Energy Principle: A Hypothesis for Intelligence
arXiv 2607.24287 · 2026-07-28微电极阵列上的气泡生长
Bubble growth on arrays of micro-electrodes
arXiv 2607.24286 · 2026-07-28风对具有群体防御和额外食物策略的捕食者-猎物动态的影响
Effect of wind on prey-predator dynamics with group defense and additional food strategy
arXiv 2607.24285 · 2026-07-28从 M 矮星 LAMOST J1332+5057 定位恒星耀斑
Locate a stellar flare from the M dwarf LAMOST J1332+5057
arXiv 2607.24284 · 2026-07-28用于构建信道知识图的信道测量位置选择：随机场理论分析
Where to Perform Channel Measurements for CKM Construction: A Random Field Theory Analysis
arXiv 2607.24283 · 2026-07-28作为N盒和量子鸽巢悖论资源的相干性
Coherence as a resource for $N$-box and quantum pigeonhole paradoxes
arXiv 2607.24282 · 2026-07-28Z/2调和1-形式的卡拉比手术
Calabi surgery for Z/2 harmonic 1-forms
arXiv 2607.24281 · 2026-07-28从专有模型到开源模型：通过智能体搜索中的多智能体协议蒸馏弥合分布差距
From Proprietary to Open-Source: Bridging the Distribution Gap via Multi-Agent Protocol Distillation in Agentic Search
arXiv 2607.24280 · 2026-07-28用于卷对卷狭缝模涂覆有机光伏活性层在线质量监测的高光谱成像
Hyperspectral imaging for inline quality monitoring of roll-to-roll slot-die coated organic photovoltaic active layers
arXiv 2607.24278 · 2026-07-28κ变形非对易时空中的修正 Mukhanov-Sasaki 方程与原初微扰
Modified Mukhanov-Sasaki equation and primordial perturbations in $κ$-deformed non-commutative space-time
arXiv 2607.24277 · 2026-07-28分词器代价：量化并解释印度语言子词分词的跨语言成本
The Tokenizer Tax: Quantifying and Explaining the Cross-Lingual Cost of Subword Tokenization for Indian Languages
arXiv 2607.24276 · 2026-07-28
