2026-09论文中文摘要
IChart2Code：面向交互式图表代码生成的多模态大语言模型基准测试
IChart2Code: Benchmarking Multimodal Large Language Models for Interactive Chart Code Generation
arXiv 2609.32413 · 2026-09-29中子星并合对应体的记忆效应
Memory effect from neutron star merger counterparts
arXiv 2609.32412 · 2026-09-29面向恢复的神经似然符号蒸馏
Recovery-Directed Symbolic Distillation of Neural Likelihoods
arXiv 2609.32409 · 2026-09-29巴萨语自动语音识别：一种低资源方法
Automatic Speech Recognition for the Basaà Language: A Low-Resource Approach
arXiv 2609.32408 · 2026-09-29打开LLM裁判：恢复最终裁决之外的偏好信号
Opening LLM Judges: Recovering Preference Signals Beyond the Final Verdict
arXiv 2609.32407 · 2026-09-29一种保结构的LOBPCG方法，用于计算若干最大的广义奇异三元组
A Structure-Preserving LOBPCG Method for Computing Several of the Largest Generalized Singular Triplets
arXiv 2609.32406 · 2026-09-29面向密集事件视觉的片上脉冲神经网络训练
Toward On-Chip Training of Spiking Neural Networks for Dense Event-Based Vision
arXiv 2609.32405 · 2026-09-29二维光滑强多凸泛函的奇异极小元
A singular minimizer of a smooth strongly polyconvex functional in two dimensions
arXiv 2609.32404 · 2026-09-29混合图的无环定向
Acyclic orientations of mixed graphs
arXiv 2609.32403 · 2026-09-29通过 $p$-Bergman 核的多次调和函数刻画及一个 Forelli 型结果
A Characterization of Pluriharmonic Functions via $p$-Bergman Kernels and A Forelli-type Result
arXiv 2609.32402 · 2026-09-29共享世界，私有心智：面向长篇写作的结构化记忆作为世界构建
Shared Worlds, Private Minds: Structured Memory for Long-Form Writing as World Creation
arXiv 2609.32401 · 2026-09-29SkillDRE：通过预执行和运行时反馈进行智能体技能的双阶段红队演化
SkillDRE: Dual-Stage Red-Team Evolution of Agent Skills via Pre-Execution and Runtime Feedback
arXiv 2609.32400 · 2026-09-29Endo-TSR：用于内窥镜重建的外观与运动的时域频谱建模
Endo-TSR: Temporal Spectral Modeling of Appearance and Motion for Endoscopic Reconstruction
arXiv 2609.32399 · 2026-09-29形式服从功能：基于副本专家机制的混合专家模型中的分布正交化
Function Over Form: Distributional Orthogonalization in Mixture-of-Experts with Replica Expert Mechanism
arXiv 2609.32398 · 2026-09-29SinBrief：一种用于僧伽罗语法律文档的混合式抽象文本摘要框架
SinBrief: A Hybrid Framework for Abstractive Text Summarisation of Sinhala Legal Documents
arXiv 2609.32397 · 2026-09-29记忆即缓存：通过构造实现精确的上下文复用与删除
Memory as a cache: Exact context reuse and deletion by construction
arXiv 2609.32395 · 2026-09-29超越脚本化搜索：通过智能体黑盒优化实现样本高效的奖励发现
Beyond Scripted Search: Sample-Efficient Reward Discovery via Agentic Black-box Optimization
arXiv 2609.32394 · 2026-09-29奖励黑客与智能体遏制失败：基于2026年Hugging Face事件的蒙特卡洛研究
Reward Hacking and Agent Containment Failure: A Monte Carlo Study Based on the 2026 Hugging Face Incident
arXiv 2609.32390 · 2026-09-29RefCompose：基于LoRA条件扩散的多参考图像生成
RefCompose: Multi-Reference Image Generation via LoRA-Conditioned Diffusion
arXiv 2609.32389 · 2026-09-29用于秩23的$3\ imes3$矩阵乘法的51次加法交替基核
A 51-Addition Alternative-Basis Kernel for Rank-23 $3\times3$ Matrix Multiplication
arXiv 2609.32388 · 2026-09-29关于带分数阶泊松核的 Littlewood-Paley $g_\alpha$ 函数的 $L^p$ 界
On $L^p$ bounds for the Littlewood-Paley $g_α$ function with fractional poisson kernel
arXiv 2609.32387 · 2026-09-29使用机器学习研究空腹血糖的预测因子：昼夜节律时间与年龄交互作用的见解
Using Machine Learning to Investigate Predictors of Fasting Blood Glucose: Insights into Circadian Timing and Age Interactions
arXiv 2609.32386 · 2026-09-29DashAct：交互式仪表板分析中GUI代理的渐进式诊断基准
DashAct: A Progressive Diagnostic Benchmark for GUI Agents in Interactive Dashboard Analysis
arXiv 2609.32385 · 2026-09-29TimeES：基于演化谱的概率与确定性时间序列预测
TimeES: Probabilistic and Deterministic Time Series Forecasting via Evolutionary Spectra
arXiv 2609.32384 · 2026-09-29关于Kähler流形正质量定理稳定性的触手示例
A tentacle example for the stability of the positive mass theorem for Kähler manifolds
arXiv 2609.32383 · 2026-09-29LLM金融智能体评估中的测量边界：固定磁带执行与多缺陷审计
Measurement Boundaries in LLM Financial Agent Evaluation: Fixed-Tape Execution and Multi-Defect Auditing
arXiv 2609.32379 · 2026-09-29AuthorityLens：通过权威视角重新思考基于LLM的智能体系统
AuthorityLens: Rethinking LLM-Based Agent Systems Through the Lens of Authority
arXiv 2609.32378 · 2026-09-29共享基础设施故障时谁买单？零液体排放、利用陷阱与印度纺织和制革集群合规成本的承担
Who Pays When Shared Infrastructure Fails? Zero Liquid Discharge, the Utilisation Trap, and the Incidence of Compliance Cost in India's Textile and Tannery Clusters
arXiv 2609.32377 · 2026-09-29一种端到端潜在展开方法，将少步ImageNet-$256$生成推进至FID $1.11$，无需Fréchet损失
An End-to-End Latent-Rollout Approach for Pushing Few-Step ImageNet-$256$ Generation to FID $1.11$ without Fréchet Losses
arXiv 2609.32376 · 2026-09-29GPUPHOT：一个用于高性能GPU加速测光与分布式天文数据归算的Python框架
GPUPHOT: A Python Framework for High-Performance GPU-Accelerated Photometry and Distributed Astronomical Data Reduction
arXiv 2609.32375 · 2026-09-29丢番图方程 $X^4+ Y^4=2Z^4$ 的矩阵解
Matrix Solutions to the Diophantine equation $X^4+ Y^4=2Z^4$
arXiv 2609.32374 · 2026-09-29主动子空间引导的自由形变与Sinkhorn自编码器用于参数化形状问题的降阶建模
Active Subspace-Guided Free-Form Deformation with Sinkhorn Autoencoders for Reduced-Order Modelling of Parametrised Shape Problems
arXiv 2609.32373 · 2026-09-29螺旋度控制的混合拓扑磁纹理中的霍尔输运
Helicity-Controlled Hall Transport in Hybrid Topological Magnetic Textures
arXiv 2609.32372 · 2026-09-29关于 Lie--Rinehart 代数的 Schouten--Nijenhuis 括号的一个注记
A Note on the Schouten--Nijenhuis Bracket of Lie--Rinehart Algebras
arXiv 2609.32371 · 2026-09-29辛Lie--Rinehart--Jacobi预量子化中的严格提升与上同调障碍
Strict lifts and cohomological obstructions in symplectic Lie--Rinehart--Jacobi prequantization
arXiv 2609.32370 · 2026-09-29弱相依估计方程的两点依赖野生自助法
Two-Point Dependent Wild Bootstrap for Weakly Dependent Estimating Equations
arXiv 2609.32369 · 2026-09-29从质量到能量-动量：能量-质量关系的场论视角
From Mass to Energy-Momentum: The Field-Theoretic Perspective on the Energy-Mass Relation
arXiv 2609.32368 · 2026-09-29STRIDE：基于增量动力学与演化的状态转移表示
STRIDE: State-Transition Representation via Increment Dynamics and Evolution
arXiv 2609.32367 · 2026-09-29第一类完全椭圆积分的端点绝对单调性
Endpoint Absolute Monotonicity for the Complete Elliptic Integral of the First Kind
arXiv 2609.32366 · 2026-09-29图记忆：基于Dirichlet能量的谱联想记忆
Graph Memory: Spectral Associative Memory via Dirichlet Energy
arXiv 2609.32365 · 2026-09-29二次NLS系统在$\mathbb{R}^{5}$中无质量共振或径向对称的散射
Scattering for the quadratic NLS system in $\mathbb{R}^{5}$ without mass resonance or radial symmetry
arXiv 2609.32364 · 2026-09-29DiffPTS：重新思考用于概率时间序列预测的扩散ELBO
DiffPTS: Rethinking Diffusion ELBO for Probabilistic Time Series Forecasting
arXiv 2609.32363 · 2026-09-29StegGNN：学习图像隐写的图形表示
StegGNN: Learning Graphical Representation for Image Steganography
arXiv 2609.32362 · 2026-09-29多智能体辩论蒸馏中认知可靠性的黑盒审计
Black-Box Auditing of Epistemic Reliability in Multi-Agent Debate Distillation
arXiv 2609.32361 · 2026-09-29可编辑地图条件下的轨迹生成用于人类移动性模拟
Editable Map-Conditioned Trajectory Generation for Human Mobility Simulation
arXiv 2609.32360 · 2026-09-29机器学习在不规则和各向异性网格上SUPG方法参数计算中的应用
Application of Machine Learning to the Computation of Parameters in the SUPG Method for Unstructured and Anisotropic Meshes
arXiv 2609.32358 · 2026-09-29布尔约束满足问题的稳定性二分法
Stability Dichotomies for Boolean Constraint Satisfaction Problems
arXiv 2609.32357 · 2026-09-29干预梯度与检测激活：语言模型中的目标特征学习
Gradients for Interventions and Activations for Detection: Targeted Feature Learning in Language Models
arXiv 2609.32355 · 2026-09-29人机协作的主动运动规划
Proactive Motion Planning for Human-Robot Cooperation
arXiv 2609.32354 · 2026-09-29从轨迹到基于环境的偏好：通过交互元素图进行Web PRM的过程偏好合成
From Trajectories to Grounded Preferences: Process Preference Synthesis via Interaction Element Graphs for Web PRMs
arXiv 2609.32351 · 2026-09-29
