2026-09论文中文摘要
从意图到算法：传输网络的验证式算法发现
From Intents to Algorithms: Verified Algorithm Discovery for Transport Networks
arXiv 2609.27386 · 2026-09-24预测工作流基准：评估语言模型在预算化预测工具下的决策能力
Forecast Workflow Bench: Evaluating Language-Model Decisions with Budgeted Forecast Tools
arXiv 2609.27385 · 2026-09-24随机分数阶亥姆霍兹方程反源问题的稳定性
Stability for the inverse source problem of the stochastic fractional Helmholtz equation
arXiv 2609.27383 · 2026-09-24当纠缠下界揭示差异：音频理解模型中的人口统计公平性审计与修复
When Entanglement Lower-Bounds Disparity: Auditing and Repairing Demographic Fairness in Audio Understanding Models
arXiv 2609.27382 · 2026-09-24CoPRE：提高低成本机械臂本体感觉接触检测的灵敏度
CoPRE: Improving Sensitivity in Proprioceptive Contact Detection for Low-Cost Robot Arms
arXiv 2609.27381 · 2026-09-24星系晕中具有径向压力的黑洞的准正则模红移
Quasinormal-mode redshifts of black holes in galactic halos with radial pressures
arXiv 2609.27379 · 2026-09-24心理声学校准的潜在平滑用于全双工语音到语音对话模型的对抗鲁棒性
Psychoacoustically Aligned Latent Smoothing for Adversarial Robustness of Full-Duplex Speech-to-Speech Dialogue Models
arXiv 2609.27378 · 2026-09-24稳定流形中的轨道与范畴熵：有限 Albanese 态射簇的应用
Orbits in stability manifolds and categorical entropy: applications to varieties with finite Albanese morphisms
arXiv 2609.27377 · 2026-09-24跨语言法律问答：越南劳动法的检索、翻译与验证器引导修正
Cross-Lingual Legal QA for Vietnamese Labour Law: Retrieval, Translation, and Verifier-Guided Correction
arXiv 2609.27376 · 2026-09-24探测 $^{12}$C+$^{12}$C 聚变反应：通过 $^{12}$C($^{14}$N,$αd$)$^{20}$Ne 准自由反应中的零度旁观者测量
Probing the $^{12}$C+$^{12}$C fusion reaction via zero-degree spectator measurement in the $^{12}$C($^{14}$N,$αd$)$^{20}$Ne quasi-free reaction
arXiv 2609.27375 · 2026-09-24规划测试时扩展与协调推理路径
Planned Test-Time Scaling with Coordinated Reasoning Paths
arXiv 2609.27374 · 2026-09-24既不沉默也不重叠是失败：全双工口语对话模型中基于意图的轮流说话评估
Neither Silence nor Overlap Is Failure: Intent-Conditioned Evaluation of Turn-Taking in Full-Duplex Spoken Dialogue Models
arXiv 2609.27372 · 2026-09-24ASAP：用于识别和分析AI生成图像中图像模式的视觉分析系统
ASAP: Visual Analytics for Identifying and Analyzing Image Patterns in AI-generated Images
arXiv 2609.27371 · 2026-09-24自然环境中基于几何条件的视觉地点识别
Geometry-Conditioned Visual Place Recognition in Natural Environments
arXiv 2609.27370 · 2026-09-24Spectral-NFP：用于加速WMMSE的认证低秩曲率主化方法
Spectral-NFP: Certified Low-Rank Curvature Majorization for Accelerating WMMSE
arXiv 2609.27369 · 2026-09-24理解公民大会中代表性的公众感知：一项实证研究
Understanding Human Perception of Representation in Citizens' Assemblies: An Empirical Study
arXiv 2609.27368 · 2026-09-24先封印，再采样：从 GPT-2 到 70B 的可验证 LLM 推理的分层采样证明
Seal, Then Sample: Sampled Layerwise Proofs for Verifiable LLM Inference from GPT-2 to 70B
arXiv 2609.27367 · 2026-09-24$3\times n$ 网格图的匹配复形的同调
Homology of matching complexes of $3\times n$ grid graphs
arXiv 2609.27366 · 2026-09-24锚定与扰动：通过探索注入实现惰性智能体的补救
Anchor and Perturb: Lazy Agent Remediation by Exploration Injection
arXiv 2609.27365 · 2026-09-24两条线段上加权测度的谱性
Spectrality of Weighted Measures on Two Line Segments
arXiv 2609.27364 · 2026-09-24从LiDAR地图到视觉定位：用于鲁棒点-线-面位姿估计的统一视觉关联
From LiDAR Maps to Visual Localization: Unified Visual Association for Robust Point-Line-Plane Pose Estimation
arXiv 2609.27363 · 2026-09-24基于AUC界的无异常自优化
Anomaly-Free Self-Optimization via AUC Bounds
arXiv 2609.27362 · 2026-09-24单射频链多用户语义通信：基于超表面调制
Single-RF-Chain Multiuser Semantic Communications via Metasurface Modulation
arXiv 2609.27361 · 2026-09-24关于交换概形的正规性
On the normality of the commuting scheme
arXiv 2609.27360 · 2026-09-24使用混合RAG和本地部署大语言模型自动提取处理活动记录（RoPA）
Automated Extraction of Records of Processing Activities (RoPA) Using Hybrid RAG and Locally Deployed Large Language Models
arXiv 2609.27359 · 2026-09-24基于内部质量驱动的全方位两栖运动
Omnidirectional Amphibious Locomotion via Internal Mass Actuation
arXiv 2609.27358 · 2026-09-24SAGEGAN：基于高斯嵌入的生成对抗网络风格异常检测
SAGEGAN: Style-Based Anomaly Detection with Gaussian Embeddings using Generative Adversarial Networks
arXiv 2609.27357 · 2026-09-24超越平均干扰：审计冻结CLIP区域解释中的最坏干扰特异性
Beyond Mean Foils: Auditing Worst-Foil Specificity in Frozen CLIP Region Explanations
arXiv 2609.27356 · 2026-09-24量化鲁棒的去学习：基于保留-遗忘损失景观交互的视角
Quantization-Robust Unlearning through the Lens of Retain-Forget Loss Landscapes Interaction
arXiv 2609.27355 · 2026-09-24约束驱动的上下文工程：为AI系统设计领域接口
Constraint-Driven Context Engineering: Designing Domain Interfaces for AI Systems
arXiv 2609.27354 · 2026-09-24引导引发行动：多模态网页智能体中引导-行动相互增强的离线研究
Guides That Cause Actions: An Offline Study of Guide-Action Mutual Reinforcement in Multimodal Web Agents
arXiv 2609.27353 · 2026-09-24范德华超导准晶 (Ta$_{0.7}$Nb$_{0.3}$)$_{1.6}$Te 的各向异性上临界场
Anisotropic upper critical field in the van der Waals superconducting quasicrystal (Ta$_{0.7}$Nb$_{0.3}$)$_{1.6}$Te
arXiv 2609.27352 · 2026-09-24中间极 2PBC J0800.5-4306 的轨道周期和自转周期
Orbital and spin periods of the intermediate polar 2PBC J0800.5-4306
arXiv 2609.27351 · 2026-09-24Baruah--Gogoi关于奇数和偶数上划线部分之和的猜想的一个证明
A proof of the Baruah--Gogoi conjecture on sums of odd and even overlined parts
arXiv 2609.27350 · 2026-09-24MolDesignBench：评估基于LLM的场景化分子设计智能体
MolDesignBench: Evaluating LLM-based Agent for Scenario-grounded Molecular Design
arXiv 2609.27349 · 2026-09-24算子复合信赖域方法的相对原始-对偶间隙证书
Relative Primal--Dual Gap Certificates for Operator-Composite Trust-Region Methods
arXiv 2609.27348 · 2026-09-24THz驱动的量子顺电体SrTiO3中的量子离子磁性
THz-Driven Quantum Ionic Magnetism in a Quantum Paraelectric SrTiO3
arXiv 2609.27347 · 2026-09-24反射感知推理用于非视距行人定位
Reflection-Aware Reasoning for Non-Line-of-Sight Pedestrian Localization
arXiv 2609.27346 · 2026-09-24一般非线性随机系统的高概率采样数据镇定
High-Probability Sampled-Data Stabilization of General Nonlinear Stochastic Systems
arXiv 2609.27345 · 2026-09-24分子光谱学的量子信道判别极限
Limits for molecular spectroscopy from quantum channel discrimination
arXiv 2609.27344 · 2026-09-24基于混合整数二阶锥规划的TDOA目标跟踪动态传感器配对
Dynamic Sensor Pairing for TDOA-Based Target Tracking via mixed-integer Second-Order Cone Programming
arXiv 2609.27343 · 2026-09-24具有奇异测度的退化Monge-Ampère方程的唯一性与尖锐边界估计
Uniqueness and sharp boundary estimates for degenerate Monge-Ampère equations with singular measures
arXiv 2609.27341 · 2026-09-24基于MCP的LLM驱动机器人导航中的空间与语义推理
Spatial and Semantic Reasoning for LLM-Driven Robot Navigation via MCP
arXiv 2609.27340 · 2026-09-24各向异性介电雪人粒子实现对光子钩的改进控制
Anisometric Dielectric Snowman-Particle Provides Improved Control Over Photonic Hook
arXiv 2609.27339 · 2026-09-24DUGM-R：面向学习型局部导航的不确定性感知动态栅格地图与风险触发恢复
DUGM-R: Uncertainty-Aware Dynamic Grid Mapping and Risk-Triggered Recovery for Learned Local Navigation
arXiv 2609.27338 · 2026-09-24利用LLM演化可检查的O-RAN切片xApps
Evolving Inspectable O-RAN Slicing xApps with LLMs
arXiv 2609.27337 · 2026-09-24CART：大型语言模型的闭环自适应红队测试
CART: Closed-Loop Adaptive Red Teaming for Large Language Models
arXiv 2609.27336 · 2026-09-24吸收连续时间马尔可夫链中的多项式衰减
Polynomial Decay in Absorbing Continuous-Time Markov Chains
arXiv 2609.27335 · 2026-09-24即时记忆：为LLM智能体学习策展任务自适应记忆
Just-in-Time Memory: Learning to Curate Task-Adaptive Memory for LLM Agents
arXiv 2609.27334 · 2026-09-24对齐惯性：通过策略覆盖阻力审计训练数据影响的持久性
Alignment Inertia: Auditing the Durability of Training Data Influence Through Policy Override Resistance
arXiv 2609.27333 · 2026-09-24
