2026-09论文中文摘要
用于机器遗忘的因果路由
Causal Routing for Unlearning
arXiv 2609.34475 · 2026-09-29深度核套期保值
Deep kernel hedging
arXiv 2609.34474 · 2026-09-29从噪声遥测到可操作警告：工业集群中的GPU故障预测
From Noisy Telemetry to Actionable Warnings: GPU Failure Prediction in Industrial Clusters
arXiv 2609.34473 · 2026-09-29VL-AcneSeg：一种用于区域感知痤疮病变分割的视觉-语言框架
VL-AcneSeg: A Vision-Language Framework for Region-Aware Acne Lesion Segmentation
arXiv 2609.34472 · 2026-09-29三个标准差足够而一个不足
Three Standard Deviations Suffice While One Does Not
arXiv 2609.34471 · 2026-09-29可交互世界的精确编辑与灵活引用
Precise Editing and Flexible Referencing for Interactable Worlds
arXiv 2609.34470 · 2026-09-29最后一英里是文件：面向保留感知的办公编辑的OfficeEditBench基准
The Last Mile Is the File: OfficeEditBench for Preservation-Aware Office Editing
arXiv 2609.34469 · 2026-09-29界面生长与种群动力学随机模型中的涨落与多重分形
Fluctuations and multifractality in stochastic models of interface growth and population dynamics
arXiv 2609.34468 · 2026-09-29PhysioTRACE：面向生理基础模型的来源感知压力测试
PhysioTRACE: Provenance-Aware Stress Tests for Physiological Foundation Models
arXiv 2609.34466 · 2026-09-29Plectic Lie代数在Hilbert模簇上的作用
Plectic Lie Algebra Action on Hilbert modular varieties
arXiv 2609.34465 · 2026-09-29从荧光显微镜图像自动生成专家级神经元分割掩膜，用于相衬图像的非侵入式深度学习分析
Automatic Generation of Expert-Level Neuron Segmentation Masks from Fluorescence Microscopy Images for Non-Invasive Deep Learning Analysis of Phase-Contrast Images
arXiv 2609.34464 · 2026-09-29CoDeL：针对基于LLM的智能体中间接提示注入的协同进化防御
CoDeL: Co-Evolutionary Defense against Indirect Prompt Injection in LLM-based Agents
arXiv 2609.34463 · 2026-09-29行列式复杂度的二次下界
A Quadratic Lower Bound on Determinantal Complexity
arXiv 2609.34462 · 2026-09-29CharDuplex：构建角色一致的全双工口语对话模型
CharDuplex: Building Character-Consistent Full-Duplex Spoken Dialogue Models
arXiv 2609.34461 · 2026-09-29结构化知识何时有助于神经定理证明？
When Does Structured Knowledge Help Neural Theorem Proving?
arXiv 2609.34460 · 2026-09-29逃离局部视角：发现潜在概念以实现可解释的多智能体强化学习
Escaping Local Views: Discovering Latent Concepts for Interpretable Multi-Agent Reinforcement Learning
arXiv 2609.34459 · 2026-09-29理解泛化需要通用归纳
Understanding Generalization Requires Universal Induction
arXiv 2609.34458 · 2026-09-29打破Windows恶意软件检测：问题空间对抗鲁棒性的全面评估
Breaking Windows Malware Detection: A Comprehensive Evaluation of Problem-Space Adversarial Robustness
arXiv 2609.34456 · 2026-09-29RGDT-Bench：面向规则治理决策及其理由的LLM推理基准
RGDT-Bench: Benchmarking LLM Reasoning for Rule-Governed Decisions and Their Justifications
arXiv 2609.34455 · 2026-09-29当言语不足时：言语化推理与隐藏特征在LLM置信度估计中的迭代协同
When Words Fall Short: Iterative Synergy Between Verbalized Reasoning and Hidden Features for LLM Confidence Estimation
arXiv 2609.34454 · 2026-09-29SPACE-LoRA：为持续学习分配激活子空间保护
SPACE-LoRA: Allocating Activation-Subspace Protection for Continual Learning
arXiv 2609.34453 · 2026-09-29Coarse-to-Fine宏单元布局：基于进化搜索与关键宏单元调优
Coarse-to-Fine Macro Placement via Evolutionary Search and Critical Macro Tuning
arXiv 2609.34452 · 2026-09-29将全切片图像建模为动态肿瘤微环境场
Modeling Whole-Slide Images as Dynamic Tumor Microenvironment Fields
arXiv 2609.34451 · 2026-09-29ReproBench：基准测试LLM智能体从零开始复现漏洞
ReproBench: Benchmarking LLM Agents on Reproducing Vulnerability From Scratch
arXiv 2609.34450 · 2026-09-29CORTEX：在稠密语言模型中学习共享与特化
CORTEX: Learning to Share and Specialize in Dense Language Models
arXiv 2609.34449 · 2026-09-29WiFi反向散射用于绿色物联网：概念、研究趋势与实际挑战
WiFi Backscatter for Green Internet of Things: Concepts, Research Trends, and Practical Challenges
arXiv 2609.34448 · 2026-09-29活在先验上：逆问题的似然分数近似
Livin' on a Prior: Likelihood Score Approximation for Inverse Problems
arXiv 2609.34446 · 2026-09-29复值信号的近端友好对数幅度先验
Prox-Friendly Log-Magnitude Prior on Complex-Valued Signal
arXiv 2609.34445 · 2026-09-29多智能体回声室背后的社会电路
Social Circuits behind Multi-agent Echo Chambers
arXiv 2609.34444 · 2026-09-29带电 Simpson-Visser-AdS 时空：QPO 与观测约束
Charged Simpson-Visser-AdS spacetime: QPOs and observational constraints
arXiv 2609.34443 · 2026-09-29让大语言模型真正遗忘：通过搜索、选择与切断知识路径实现深度遗忘
Making LLMs Truly Forget: Deep Unlearning by Searching, Selecting, and Severing Knowledge Paths
arXiv 2609.34442 · 2026-09-29$\triangle$-凸性中Carathéodory数与交换数的紧界
Tight bounds on the Carathéodory and exchange numbers in $\triangle$-convexity
arXiv 2609.34441 · 2026-09-29当分数成为目标：重新思考自动驾驶中的指标有效性
When the Score Becomes the Target: Rethinking Metric Validity in Autonomous Driving
arXiv 2609.34440 · 2026-09-29临床轨迹对齐用于医学视觉-语言预训练
Clinical Trajectory Alignment for Medical Vision-Language Pre-training
arXiv 2609.34439 · 2026-09-29通过提问来记忆：面向LLM智能体的检索诱导记忆演化
Remember by Asking: Retrieval-Induced Memory Evolution for LLM Agents
arXiv 2609.34438 · 2026-09-29高维局部类域论
Higher dimensional local class field theory
arXiv 2609.34437 · 2026-09-29具有小偏导数空间的多项式的命中集
Hitting Sets for Polynomials with Small Partial Derivative Spaces
arXiv 2609.34436 · 2026-09-29关于具有全纯不变Kähler--Berwald度量的有界对称域上的Schwarz引理
Schwarz lemma on bounded symmetric domains endowed with holomorphic invariant Kähler--Berwald metrics
arXiv 2609.34435 · 2026-09-29phi-半可收缩性
phi-semi contractible
arXiv 2609.34434 · 2026-09-29多模态干预轨迹的可容许扩散
Admissible Diffusion for Multimodal Interventional Trajectories
arXiv 2609.34433 · 2026-09-29语义、工作流与基础设施：理解生产规模下的智能体服务
Semantics, Workflows, and Infrastructure: Understanding Agent Serving at Production Scale
arXiv 2609.34432 · 2026-09-29HUMAN-TCI：面向文本到运动检索的以躯干为中心的层次化多流运动感知网络
HUMAN-TCI: Hierarchical Multi-Stream Motion-Aware Network with Torso-Centered Interaction for Text-to-Motion Retrieval
arXiv 2609.34430 · 2026-09-29三次曲面的对数 Calabi-Yau 模空间的壁穿越
Log Calabi-Yau wall crossing for moduli of cubic surfaces
arXiv 2609.34429 · 2026-09-29AgentHop：面向智能体多跳科学问答的诊断基准
AgentHop: A Diagnostic Benchmark for Agentic Multi-Hop Scientific Question Answering
arXiv 2609.34428 · 2026-09-29LLMs作为自适应元求解器：面向工业规模优化的策略多样化强化学习
LLMs as Adaptive Meta-Solvers: Strategy-Diverse RL for Industrial-Scale Optimization
arXiv 2609.34427 · 2026-09-29口语文档的零样本线索引导主题分割
Zero-Shot Cue-Grounded Topic Segmentation of Spoken Documents
arXiv 2609.34425 · 2026-09-29固定秩阈值筛选中精确证书的构造-复用权衡
Construction-Reuse Trade-offs for Exact Certificates in Fixed-Rank Threshold Screening
arXiv 2609.34424 · 2026-09-29关于区间遗憾与动态遗憾之间关系的探讨
On the Relation Between Interval Regret and Dynamic Regret
arXiv 2609.34423 · 2026-09-29发现 SNR HB 9 内壳层中的 Kolmogorov 型湍流
Discovery of Kolmogorov-like turbulence in the inner shell of SNR HB 9
arXiv 2609.34421 · 2026-09-29MegaGraph：面向大规模图Transformer的高效训练与自动化混合并行
MegaGraph: Towards Efficient Training of Large-Scale Graph Transformers with Automated Hybrid Parallelism
arXiv 2609.34420 · 2026-09-29
