机器学习学术速递[9.16]

点击阅读原文访问arxivdaily.com，涵盖CS|物理|数学|经济|统计|金融|生物|电气领域，更有搜索、收藏等功能！

cs.LG 方向，今日共计269篇

大模型相关(28篇)

【1】AMQ: Enabling AutoML for Mixed-precision Weight-Only Quantization of Large Language Models
标题：AMQ：启用AutoML以实现大型语言模型的混合精度仅加权量化
链接：https://arxiv.org/abs/2509.12019

【2】MillStone: How Open-Minded Are LLMs?
标题：MillStone：法学硕士的思想开放程度如何？
链接：https://arxiv.org/abs/2509.11967

【3】Bridging Vision Language Models and Symbolic Grounding for Video Question Answering
标题：视频问题解答的视觉语言模型和符号基础的桥梁
链接：https://arxiv.org/abs/2509.11862

【4】Collapse of Irrelevant Representations (CIR) Ensures Robust and Non-Disruptive LLM Unlearning
标题：不相关表示（CIR）的崩溃确保稳健且非破坏性的LLM取消学习
链接：https://arxiv.org/abs/2509.11816

【5】PeruMedQA: Benchmarking Large Language Models (LLMs) on Peruvian Medical Exams - Dataset Construction and Evaluation
标题：PeruMedQA：秘鲁医学考试中的大型语言模型（LLM）基准-数据集构建和评估
链接：https://arxiv.org/abs/2509.11517

【6】Cross-Platform Scaling of Vision-Language-Action Models from Edge to Cloud GPUs
标题：从边缘到云图形处理器的视觉-语言-动作模型的跨平台扩展
链接：https://arxiv.org/abs/2509.11480

【7】Enhancing Generalization in Vision-Language-Action Models by Preserving Pretrained Representations
标题：通过保留预训练的表示来增强视觉-语言-动作模型的概括
链接：https://arxiv.org/abs/2509.11417

【8】Intelligent Reservoir Decision Support: An Integrated Framework Combining Large Language Models, Advanced Prompt Engineering, and Multimodal Data Fusion for Real-Time Petroleum Operations
标题：智能储层决策支持：一个集成框架，结合大型语言模型、高级即时工程和多模式数据融合，用于实时石油操作
链接：https://arxiv.org/abs/2509.11376

【9】PersonaX: Multimodal Datasets with LLM-Inferred Behavior Traits
标题：PersonaX：具有LLM推断行为特征的多峰数据集
链接 ：https://arxiv.org/abs/2509.11362

【10】MatQnA: A Benchmark Dataset for Multi-modal Large Language Models in Materials Characterization and Analysis
标题：MatQnA：材料特征化和分析中多模式大型语言模型的基准数据集
链接：https://arxiv.org/abs/2509.11335

【11】AQUA: Attention via QUery mAgnitudes for Memory and Compute Efficient Inference in LLMs
标题：AQUA：通过QUery mAgnspel关注LLM中的记忆和计算高效推理
链接：https://arxiv.org/abs/2509.11155

【12】Fluid Language Model Benchmarking
标题：流体语言模型基准
链接：https://arxiv.org/abs/2509.11106

【13】The Psychogenic Machine: Simulating AI Psychosis, Delusion Reinforcement and Harm Enablement in Large Language Models
标题：心因机器：模拟大型语言模型中的人工智能精神病、错觉强化和伤害启用
链接：https://arxiv.org/abs/2509.10970

【14】HalluField: Detecting LLM Hallucinations via Field-Theoretic Modeling
标题：HalluField：通过场论建模检测LLM幻觉
链接：https://arxiv.org/abs/2509.10753

【15】PolyTruth: Multilingual Disinformation Detection using Transformer-Based Language Models
标题：PolyTruth：使用基于转换器的语言模型的多语言虚假信息检测
链接：https://arxiv.org/abs/2509.10737

【16】Using LLMs for Late Multimodal Sensor Fusion for Activity Recognition
标题：使用LLM进行后期多模式传感器融合以进行活动识别
链接：https://arxiv.org/abs/2509.10729

【17】CrunchLLM: Multitask LLMs for Structured Business Reasoning and Outcome Prediction
标题：CrunchLLM：用于结构化业务推理和结果预测的多任务LLM
链接：https://arxiv.org/abs/2509.10698

【18】Struct-Bench: A Benchmark for Differentially Private Structured Text Generation
标题：Struct-Bench：差异私有结构文本生成的基准
链接：https://arxiv.org/abs/2509.10696

【19】LLM in the Middle: A Systematic Review of Threats and Mitigations to Real-World LLM-based Systems
标题：中间的法学硕士：对现实世界基于法学硕士的系统的威胁和缓解措施的系统性审查
链接：https://arxiv.org/abs/2509.10682

【20】Test-Time Warmup for Multimodal Large Language Models
标题：多模式大型语言模型的测试时热身
链接：https://arxiv.org/abs/2509.10641

【21】SME-TEAM: Leveraging Trust and Ethics for Secure and Responsible Use of AI and LLMs in SMEs
标题：中小企业团队：利用信任和道德，在中小企业中安全、负责任地使用人工智能和LLM
链接：https://arxiv.org/abs/2509.10594

【22】Uncovering the Vulnerability of Large Language Models in the Financial Domain via Risk Concealment
标题：通过风险隐藏揭示金融领域大型语言模型的脆弱性
链接：https://arxiv.org/abs/2509.10546

【23】Holographic Knowledge Manifolds: A Novel Pipeline for Continual Learning Without Catastrophic Forgetting in Large Language Models
标题：全息知识库：大型语言模型中不发生灾难性遗忘的连续学习的新型管道
链接：https://arxiv.org/abs/2509.10518

【24】The Anti-Ouroboros Effect: Emergent Resilience in Large Language Models from Recursive Selective Feedback
标题：反噬身蛇效应：大型语言模型中来自循环选择性反馈的紧急弹性
链接：https://arxiv.org/abs/2509.10509

【25】The LLM as a Network Operator: A Vision for Generative AI in the 6G Radio Access Network
标题：LLM作为网络运营商：6G无线电接入网络中生成人工智能的愿景
链接：https://arxiv.org/abs/2509.10478

【26】When marine radar target detection meets pretrained large language models
标题：当海洋雷达目标检测满足预训练的大型语言模型时
链接：https://arxiv.org/abs/2509.12110

【27】Trading-R1: Financial Trading with LLM Reasoning via Reinforcement Learning
标题：Trading-R1：通过强化学习进行LLM推理的金融交易
链接：https://arxiv.org/abs/2509.11420

【28】Predictable Compression Failures: Why Language Models Actually Hallucinate
标题：可预测的压缩失败：为什么语言模型实际上会产生幻觉
链接：https://arxiv.org/abs/2509.11208

Graph相关(图学习|图神经网络|图优化等)(18篇)

【1】Learning Contact Dynamics for Control with Action-conditioned Face Interaction Graph Networks
标题：使用条件条件面部交互图网络学习接触动力学以实现控制
链接：https://arxiv.org/abs/2509.12151

【2】Draw a Portrait of Your Graph Data: An Instance-Level Profiling Framework for Graph-Structured Data
标题：绘制图形数据的肖像：图形结构数据的实例级分析框架
链接：https://arxiv.org/abs/2509.12094

【3】Travel Time and Weather-Aware Traffic Forecasting in a Conformal Graph Neural Network Framework
标题：保形图神经网络框架中的旅行时间和天气感知交通预测
链接：https://arxiv.org/abs/2509.12043

【4】Visualization and Analysis of the Loss Landscape in Graph Neural Networks
标题：图神经网络损失格局的可视化与分析
链接：https://arxiv.org/abs/2509.11792

【5】Drug Repurposing Using Deep Embedded Clustering and Graph Neural Networks
标题：使用深度嵌入式集群和图神经网络进行药物再利用
链接：https://arxiv.org/abs/2509.11493

【6】Derivative-informed Graph Convolutional Autoencoder with Phase Classification for the Lifshitz-Petrich Model
标题：Lifshitz-Petrich模型的具有相分类的导信息图卷积自动编码器
链接：https://arxiv.org/abs/2509.11293

【7】BIGNet: Pretrained Graph Neural Network for Embedding Semantic, Spatial, and Topological Data in BIM Models
标题：BIGNet：预训练的图神经网络，用于在BMI模型中嵌入语义、空间和布局数据
链接：https://arxiv.org/abs/2509.11104

【8】Factor Graph Optimization for Leak Localization in Water Distribution Networks
标题：供水管网泄漏定位的因子图优化
链接：https://arxiv.org/abs/2509.10982

【9】GTHNA: Local-global Graph Transformer with Memory Reconstruction for Holistic Node Anomaly Evaluation
标题 ：GTHNA：具有内存重建的局部-全局图Transformer，用于整体节点异常评估
链接：https://arxiv.org/abs/2509.10869

【10】CogGNN: Cognitive Graph Neural Networks in Generative Connectomics
标题：CogGNN：生成连接组学中的认知图神经网络
链接：https://arxiv.org/abs/2509.10864

【11】Verifying Computational Graphs in Production-Grade Distributed Machine Learning Frameworks
标题：生产级分布式机器学习框架中的调试计算图
链接：https://arxiv.org/abs/2509.10694

【12】M4GN: Mesh-based Multi-segment Hierarchical Graph Network for Dynamic Simulations
标题：M4 GN：用于动态模拟的基于网格的多段分层图网络
链接：https://arxiv.org/abs/2509.10659

【13】STM-Graph: A Python Framework for Spatio-Temporal Mapping and Graph Neural Network Predictions
标题：STM-Shape：用于时空映射和图形神经网络预测的Python框架
链接：https://arxiv.org/abs/2509.10528

【14】Resource-Aware Neural Network Pruning Using Graph-based Reinforcement Learning
标题：使用基于图的强化学习的资源感知神经网络修剪
链接：https://arxiv.org/abs/2509.10526

【15】Quantum Graph Attention Networks: Trainable Quantum Encoders for Inductive Graph Learning
标题：量子图注意力网络：用于归纳图学习的可训练量子编码器
链接：https://arxiv.org/abs/2509.11390

【16】On the Impact of Downstream Tasks on Sampling and Reconstructing Noisy Graph Signals
标题：下游任务对采样和重建有噪图信号的影响
链接：https://arxiv.org/abs/2509.10874

【17】Assessing the Limits of Graph Neural Networks for Vapor-Liquid Equilibrium Prediction: A Cryogenic Mixture Case Study
标题：评估图神经网络用于气液平衡预测的局限性：低温混合物案例研究
链接：https://arxiv.org/abs/2509.10565

【18】FireGNN: Neuro-Symbolic Graph Neural Networks with Trainable Fuzzy Rules for Interpretable Medical Image Classification
标题：FireGNN：具有可训练模糊规则的神经符号图神经网络，用于可解释医学图像分类
链接：https://arxiv.org/abs/2509.10510

Transformer(9篇)

【1】Dynamic Relational Priming Improves Transformer in Multivariate Time Series
标题：动态关系启动改进多元时间序列中的Transformer
链接：https://arxiv.org/abs/2509.12196

【2】SpeCa: Accelerating Diffusion Transformers with Speculative Feature Caching
标题：SpeCa：具有推测性特征缓存的加速扩散变换器
链接：https://arxiv.org/abs/2509.11628

【3】Tabular Data with Class Imbalance: Predicting Electric Vehicle Crash Severity with Pretrained Transformers (TabPFN) and Mamba-Based Models
标题：具有类别不平衡的表格数据：使用预训练的Transformer（TabPFN）和基于Mamba的模型预测电动汽车碰撞严重程度
链接：https://arxiv.org/abs/2509.11449

【4】TransZero: Parallel Tree Expansion in MuZero using Transformer Networks
标题：TransZero：使用Transformer网络在MuZero中进行并行树扩展
链接：https://arxiv.org/abs/2509.11233

【5】Multi-Modal Sensing Aided mmWave Beamforming for V2V Communications with Transformers
标题：用于与Transformer进行V2V通信的多模式传感辅助毫米波束形成
链接：https://arxiv.org/abs/2509.11112

【6】GoldenTransformer: A Modular Fault Injection Framework for Transformer Robustness Research
标题：GoldenTransformer：用于Transformer稳健性研究的模块化故障注入框架
链接：https://arxiv.org/abs/2509.10790

【7】Kalman Bayesian Transformer
标题：卡尔曼贝氏Transformer
链接：https://arxiv.org/abs/2509.10695

【8】Wavelet-SARIMA-Transformer: A Hybrid Model for Rainfall Forecasting
标题：Wavelet-SARIMA-Transformer：降雨预测的混合模型
链接：https://arxiv.org/abs/2509.11903

【9】Adaptive Temporal Fusion Transformers for Cryptocurrency Price Prediction
标题：用于加密货币价格预测的自适应时间融合转换器
链接：https://arxiv.org/abs/2509.10542

GAN|对抗|攻击|生成相关(13篇)

【1】From Autoencoders to CycleGAN: Robust Unpaired Face Manipulation via Adversarial Learning
标题：从自动编码器到CycleGAN：通过对抗学习进行稳健的非配对面部操纵
链接：https://arxiv.org/abs/2509.12176

【2】LEGO: Spatial Accelerator Generation and Optimization for Tensor Applications
标题：乐高：针对张量应用的空间加速器生成和优化
链接：https://arxiv.org/abs/2509.12053

【3】DRAG: Data Reconstruction Attack using Guided Diffusion
标题：DRAG：使用引导扩散的数据重建攻击
链接：https://arxiv.org/abs/2509.11724

【4】CoachMe: Decoding Sport Elements with a Reference-Based Coaching Instruction Generation Model
标题：CoachMe：基于参考的教练指导生成模型解码运动元素
链接：https://arxiv.org/abs/2509.11698

【5】A Controllable 3D Deepfake Generation Framework with Gaussian Splatting
标题：一种基于高斯溅射的可控3D Deepfake生成框架
链接：https://arxiv.org/abs/2509.11624

【6】DARD: Dice Adversarial Robustness Distillation against Adversarial Attacks
标题：DARD：对抗性攻击的对抗鲁棒性蒸馏
链接：https://arxiv.org/abs/2509.11525

【7】On the Escaping Efficiency of Distributed Adversarial Training Algorithms
标题：分布式对抗训练算法的逃避效率研究
链接：https://arxiv.org/abs/2509.11337

【8】ToMA: Token Merge with Attention for Image Generation with Diffusion Models
标题：ToMA：关注代币合并，利用扩散模型生成图像
链接：https://arxiv.org/abs/2509.10918

【9】Robust DDoS-Attack Classification with 3D CNNs Against Adversarial Methods
标题：使用3D CNN对抗对抗性方法的稳健DDoS攻击分类
链接：https://arxiv.org/abs/2509.10543

【10】DualAlign: Generating Clinically Grounded Synthetic Data
标题：DualAlign：生成基于临床的合成数据
链接：https://arxiv.org/abs/2509.10538

【11】Semantic-guided LoRA Parameters Generation
标题：语义引导的LoRA参数生成
链接：https://arxiv.org/abs/2509.10535

【12】The 1st International Workshop on Disentangled Representation Learning for Controllable Generation (DRL4Real): Methods and Results
标题：第一届可控生成分解表示学习国际研讨会（DRL 4Real）：方法和结果
链接：https://arxiv.org/abs/2509.10463

【13】Next-Generation Reservoir Computing for Dynamical Inference
标题：用于动态推理的下一代水库计算
链接：https://arxiv.org/abs/2509.11338

半/弱/无/有监督|不确定性|主动学习(12篇)

【1】High Effort, Low Gain: Fundamental Limits of Active Learning for Linear Dynamical Systems
标题：高努力、低收益：线性动态系统主动学习的基本限制
链接：https://arxiv.org/abs/2509.11907

【2】Learning Representations in Video Game Agents with Supervised Contrastive Imitation Learning
标题：具有监督对比学习的视频游戏代理中的学习表示
链接：https://arxiv.org/abs/2509.11880

【3】Inducing Uncertainty for Test-Time Privacy
标题：引入测试时隐私的不确定性
链接：https://arxiv.org/abs/2509.11625

【4】Modality-Aware Infrared and Visible Image Fusion with Target-Aware Supervision
标题：具有目标感知监督的模式感知红外和可见光图像融合
链接：https://arxiv.org/abs/2509.11476

【5】Beyond Instance Consistency: Investigating View Diversity in Self-supervised Learning
标题：超越实例一致性：调查自我监督学习中的观点多样性
链接：https://arxiv.org/abs/2509.11344

【6】Matched-Pair Experimental Design with Active Learning
标题：具有主动学习的匹配对实验设计
链接：https://arxiv.org/abs/2509.10742

【7】Self-Supervised Goal-Reaching Results in Multi-Agent Cooperation and Exploration
标题：多智能体合作和探索中自我监督的目标达成结果
链接：https://arxiv.org/abs/2509.10656

【8】Building a General SimCLR Self-Supervised Foundation Model Across Neurological Diseases to Advance 3D Brain MRI Diagnoses
标题：构建神经系统疾病的通用Simpline自我监督基金会模型以推进3D脑部MRI诊断
链接：https://arxiv.org/abs/2509.10620

【9】Adaptive Preference Optimization with Uncertainty-aware Utility Anchor
标题：具有不确定性感知效用锚的自适应偏好优化
链接：https://arxiv.org/abs/2509.10515

【10】Parameter estimation with uncertainty quantification from continuous measurement data using neural network ensembles
标题：使用神经网络集成从连续测量数据进行不确定性量化的参数估计
链接：https://arxiv.org/abs/2509.10756

【11】HiLWS: A Human-in-the-Loop Weak Supervision Framework for Curating Clinical and Home Video Data for Neurological Assessment
标题：HiLWS：一个人在环弱监督框架，用于管理神经系统评估的临床和家庭视频数据
链接：https://arxiv.org/abs/2509.10557

【12】Data-Efficient Psychiatric Disorder Detection via Self-supervised Learning on Frequency-enhanced Brain Networks
标题：通过频率增强大脑网络上的自我监督学习进行数据高效的精神疾病检测
链接：https://arxiv.org/abs/2509.10524

迁移|Zero/Few/One-Shot|自适应(8篇)

【1】FedDAF: Federated Domain Adaptation Using Model Functional Distance
标题：Federator：使用模型功能距离的联邦域自适应
链接：https://arxiv.org/abs/2509.11819

【2】Adaptive-GraphSketch: Real-Time Edge Anomaly Detection via Multi-Layer Tensor Sketching and Temporal Decay
标题：Adaptive-GraphSketch：通过多层张量绘制和时间衰减进行实时边缘异常检测
链接：https://arxiv.org/abs/2509.11633

【3】Dynamic Adaptive Parsing of Temporal and Cross-Variable Patterns for Network State Classification
标题：用于网络状态分类的时间和交叉变量模式的动态自适应解析
链接：https://arxiv.org/abs/2509.11601

【4】Neurosymbolic AI Transfer Learning Improves Network Intrusion Detection
标题：神经符号人工智能迁移学习改进网络入侵检测
链接：https://arxiv.org/abs/2509.10850

【5】FinXplore: An Adaptive Deep Reinforcement Learning Framework for Balancing and Discovering Investment Opportunities
标题：FinXplore：用于平衡和发现投资机会的自适应深度强化学习框架
链接：https://arxiv.org/abs/2509.10531

【6】Dynamic Adaptive Shared Experts with Grouped Multi-Head Attention Mixture of Experts
标题：具有分组多头注意力专家混合的动态自适应共享专家
链接：https://arxiv.org/abs/2509.10530

【7】A Service-Oriented Adaptive Hierarchical Incentive Mechanism for Federated Learning
标题：面向服务的联邦学习自适应分层激励机制
链接：https://arxiv.org/abs/2509.10512

【8】SABR: A Stable Adaptive Bitrate Framework Using Behavior Cloning Pretraining and Reinforcement Learning Fine-Tuning
标题：SABR：使用行为克隆预训练和强化学习微调的稳定自适应比特率框架
链接：https://arxiv.org/abs/2509.10486

强化学习(7篇)

【1】$K$-Level Policy Gradients for Multi-Agent Reinforcement Learning
标题：$K$-多智能体强化学习的级别策略要素
链接：https://arxiv.org/abs/2509.12117

【2】Generalizing Behavior via Inverse Reinforcement Learning with Closed-Form Reward Centroids
标题：通过具有封闭形式奖励中心的反向强化学习来概括行为
链接：https://arxiv.org/abs/2509.12010

【3】UI-S1: Advancing GUI Automation via Semi-online Reinforcement Learning
标题：UI-S1：通过半在线强化学习推进图形界面自动化
链接：https://arxiv.org/abs/2509.11543

【4】SafeDiver: Cooperative AUV-USV Assisted Diver Communication via Multi-agent Reinforcement Learning Approach
标题：SafeDiver：通过多智能体强化学习方法进行AUV-USV合作辅助潜水员沟通
链接：https://arxiv.org/abs/2509.11508

【5】Gradient Free Deep Reinforcement Learning With TabPFN
标题：使用TabPFN的无梯度深度强化学习
链接：https://arxiv.org/abs/2509.11259

【6】Coordinated Reinforcement Learning Prefetching Architecture for Multicore Systems
标题：多核系统的协调强化学习预取架构
链接：https://arxiv.org/abs/2509.10719

【7】LogGuardQ: A Cognitive-Enhanced Reinforcement Learning Framework for Cybersecurity Anomaly Detection in Security Logs
标题：LogGuardQ：用于安全收件箱中网络安全异常检测的认知增强强化学习框架
链接：https://arxiv.org/abs/2509.10511

符号|符号学习(2篇)

【1】Neuro-Symbolic Agents with Modal Logic for Autonomous Diagnostics
标题：具有模式逻辑的神经符号代理用于自主诊断
链接：https://arxiv.org/abs/2509.11943

【2】Exploring Multi-view Symbolic Regression methods in physical sciences
标题：物理科学中的多视图符号回归方法探讨
链接：https://arxiv.org/abs/2509.10500

医学相关(4篇)

【1】Disentanglement of Biological and Technical Factors via Latent Space Rotation in Clinical Imaging Improves Disease Pattern Discovery
标题：临床成像中通过潜在空间旋转解开生物和技术因素，改善疾病模式发现
链接：https://arxiv.org/abs/2509.11436

【2】A Comparative Benchmark of Federated Learning Strategies for Mortality Prediction on Heterogeneous and Imbalanced Clinical Data
标题：用于非均匀和不平衡临床数据死亡率预测的联邦学习策略比较基准
链接：https://arxiv.org/abs/2509.10517

【3】EMeRALDS: Electronic Medical Record Driven Automated Lung Nodule Detection and Classification in Thoracic CT Images
标题：EmeRALDS：电子医疗记录驱动的胸部CT图像中的自动肺部结节检测和分类
链接：https://arxiv.org/abs/2509.11714

【4】FlowECG: Using Flow Matching to Create a More Efficient ECG Signal Generator
标题：Flow心电图：使用流量匹配创建更高效的心电图信号发生器
链接：https://arxiv.org/abs/2509.10491

蒸馏|知识提取(3篇)

【1】Stabilizing Data-Free Model Extraction
标题：稳定无数据模型提取
链接：https://arxiv.org/abs/2509.11159

【2】PHLoRA: data-free Post-hoc Low-Rank Adapter extraction from full-rank checkpoint
标题：PHLoRA：从全级别检查点提取无数据的事后低级别适配器
链接：https://arxiv.org/abs/2509.10971

【3】Predictive Free Energy Simulations Through Hierarchical Distillation of Quantum Hamiltonians
标题：通过量子Hamilton分层蒸馏进行预测自由能模拟
链接：https://arxiv.org/abs/2509.10967

推荐(3篇)

【1】Federated Recommender System with Data Valuation for E-commerce Platform
标题：电子商务平台数据评估联合推荐系统
链接：https://arxiv.org/abs/2509.11196

【2】Privacy-Preserving Personalization in Education: A Federated Recommender System for Student Performance Prediction
标题：保护隐私的教育个性化：学生表现预测的联邦推荐系统
链接：https://arxiv.org/abs/2509.10516

【3】Momentum-integrated Multi-task Stock Recommendation with Converge-based Optimization
标题：基于融合优化的动量集成多任务股票推荐
链接：https://arxiv.org/abs/2509.10461

聚类(1篇)

【1】MMM: Clustering Multivariate Longitudinal Mixed-type Data
标题：MMM：对多元纵向混合型数据进行聚集
链接：https://arxiv.org/abs/2509.12166

自动驾驶|车辆|车道检测等(1篇)

【1】Retrosynthesis Planning via Worst-path Policy Optimisation in Tree-structured MDPs
标题：基于最差路径策略优化的树型MDP逆合成规划
链接：https://arxiv.org/abs/2509.10504

联邦学习|隐私保护|加密(3篇)

【1】Accurate and Private Diagnosis of Rare Genetic Syndromes from Facial Images with Federated Deep Learning
标题：利用联邦深度学习从面部图像准确且私人地诊断罕见遗传综合症
链接：https://arxiv.org/abs/2509.10635

【2】On Using Large-Batches in Federated Learning
标题：关于在联邦学习中使用大批量学习
链接：https://arxiv.org/abs/2509.10537

【3】Cost-Free Personalization via Information-Geometric Projection in Bayesian Federated Learning
标题：通过Bayesian联邦学习中的信息几何投影实现无成本个性化
链接：https://arxiv.org/abs/2509.10132

推理|分析|理解|解释(9篇)

【1】Early Detection of Branched Broomrape (Phelipanche ramosa) Infestation in Tomato Crops Using Leaf Spectral Analysis and Machine Learning
标题：利用叶片光谱分析和机器学习早期检测番茄作物中枝蔓越菜（Volipanche ramosa）侵扰
链接：https://arxiv.org/abs/2509.12074

【2】Data-Driven Analysis of Text-Conditioned AI-Generated Music: A Case Study with Suno and Udio
标题：文本条件人工智能生成音乐的数据驱动分析：Suno和Udio的案例研究
链接：https://arxiv.org/abs/2509.11824

【3】Measuring Visual Understanding in Telecom domain: Performance Metrics for Image-to-UML conversion using VLMs
标题：电信领域中的视觉理解测量：使用VLM的图像到UML转换的性能测试
链接：https://arxiv.org/abs/2509.11667

【4】OASIS: A Deep Learning Framework for Universal Spectroscopic Analysis Driven by Novel Loss Functions
标题：OASIS：由新型损失函数驱动的通用光谱分析深度学习框架
链接：https://arxiv.org/abs/2509.11499

【5】Your Compiler is Backdooring Your Model: Understanding and Exploiting Compilation Inconsistency Vulnerabilities in Deep Learning Compilers
标题：您的编译器正在为您的模型做后门：了解和利用深度学习编译器中的编译不一致漏洞
链接：https://arxiv.org/abs/2509.11173

【6】From Predictions to Explanations: Explainable AI for Autism Diagnosis and Identification of Critical Brain Regions
标题：从预测到解释：用于自闭症诊断和关键大脑区域识别的可解释人工智能
链接：https://arxiv.org/abs/2509.10523

【7】A Differential Manifold Perspective and Universality Analysis of Continuous Attractors in Artificial Neural Networks
标题：人工神经网络中连续吸引子的差模透视和普适性分析
链接：https://arxiv.org/abs/2509.10514

【8】Branched Broomrape Detection in Tomato Farms Using Satellite Imagery and Time-Series Analysis
标题：利用卫星图像和时间序列分析检测番茄农场的枝状扫帚
链接：https://arxiv.org/abs/2509.10804

【9】Why Bonds Fail Differently? Explainable Multimodal Learning for Multi-Class Default Prediction
标题：为什么债券会失败？用于多类违约预测的可解释多模式学习
链接：https://arxiv.org/abs/2509.10802

检测相关(5篇)

【1】Improving Out-of-Domain Audio Deepfake Detection via Layer Selection and Fusion of SSL-Based Countermeasures
标题：通过基于SSL的对策的层选择和融合改进域外音频深度伪造检测
链接：https://arxiv.org/abs/2509.12003

【2】Watch Your Step: A Cost-Sensitive Framework for Accelerometer-Based Fall Detection in Real-World Streaming Scenarios
标题：注意脚下：在现实世界流媒体场景中基于加速计的跌倒检测的成本敏感框架
链接：https://arxiv.org/abs/2509.11789

【3】Detecting Model Drifts in Non-Stationary Environment Using Edit Operation Measures
标题：使用编辑操作措施检测非静止环境中的模型漂移
链接：https://arxiv.org/abs/2509.11367

【4】Agentic Username Suggestion and Multimodal Gender Detection in Online Platforms: Introducing the PNGT-26K Dataset
标题：在线平台中的显式语音建议和多模式性别检测：介绍PNGT-26 K数据集
链接：https://arxiv.org/abs/2509.11136

【5】FEDEXCHANGE: Bridging the Domain Gap in Federated Object Detection for Free
标题：FEDEXCHANGE：免费弥合联邦对象检测中的领域差距
链接：https://arxiv.org/abs/2509.10503

分类|识别(10篇)

【1】Scaling to Multimodal and Multichannel Heart Sound Classification: Fine-Tuning Wav2Vec 2.0 with Synthetic and Augmented Biosignals
标题：扩展到多模式和多通道心脏声音分类：利用合成和增强生物信号进行微调Wav2Vec 2.0
链接：https://arxiv.org/abs/2509.11606

【2】Protected Probabilistic Classification Library
标题：受保护的概率分类库
链接：https://arxiv.org/abs/2509.11267

【3】Machine Learning Framework for Audio-Based Equipment Condition Monitoring: A Comparative Study of Classification Algorithms
标题：基于音频的设备状态监控的机器学习框架：分类算法的比较研究
链接：https://arxiv.org/abs/2509.11075

【4】A Comparison of Selected Image Transformation Techniques for Malware Classification
标题：用于恶意软件分类的选定图像转换技术的比较
链接：https://arxiv.org/abs/2509.10838

【5】Least-Ambiguous Multi-Label Classifier
标题：最不模糊的多标签分类器
链接：https://arxiv.org/abs/2509.10689

【6】Interpretable neural network system identification method for two families of second-order systems based on characteristic curves
标题：基于特征曲线的两类二阶系统可解释神经网络系统识别方法
链接：https://arxiv.org/abs/2509.10632

【7】Offline Contextual Bandit with Counterfactual Sample Identification
标题：具有反事实样本识别的离线背景盗贼
链接：https://arxiv.org/abs/2509.10520

【8】Spectral and Rhythm Features for Audio Classification with Deep Convolutional Neural Networks
标题：利用深度卷积神经网络进行音频分类的频谱和节奏特征
链接：https://arxiv.org/abs/2410.06927

【9】Learning Majority-to-Minority Transformations with MMD and Triplet Loss for Imbalanced Classification
标题：利用MMD和三重损失学习多数到少数变换以实现不平衡分类
链接：https://arxiv.org/abs/2509.11511

【10】Crystal Systems Classification of Phosphate-Based Cathode Materials Using Machine Learning for Lithium-Ion Battery
标题：利用机器学习对锂离子电池磷酸盐基正极材料的晶体体系分类
链接：https://arxiv.org/abs/2509.10532

表征(4篇)

【1】Event2Vec: A Geometric Approach to Learning Composable Representations of Event Sequences
标题：Event2Vec：学习事件序列可组合表示的几何方法
链接：https://arxiv.org/abs/2509.12188

【2】GCN-TULHOR: Trajectory-User Linking Leveraging GCNs and Higher-Order Spatial Representations
标题：GCN-TULHOR：利用GCN和更高级空间表示的轨迹用户链接
链接：https://arxiv.org/abs/2509.11095

【3】SH-SAS: An Implicit Neural Representation for Complex Spherical-Harmonic Scattering Fields for 3D Synthetic Aperture Sonar
标题：SH-SAS：三维合成孔径声纳复杂球谐散射场的隐式神经表示
链接：https://arxiv.org/abs/2509.11087

【4】Contrastive Network Representation Learning
标题：对比网络表示学习
链接：https://arxiv.org/abs/2509.11316

3D|3D重建等相关(1篇)

【1】WildSmoke: Ready-to-Use Dynamic 3D Smoke Assets from a Single Video in the Wild
标题：WildSmoke：来自野外单个视频的即可使用的动态3D烟雾资源
链接：https://arxiv.org/abs/2509.11114

编码器(2篇)

【1】Lightweight Metadata-Aware Mixture-of-Experts Masked Autoencoder for Earth Observation
标题：用于地球观测的轻量级元数据感知专家混合掩蔽自动编码器
链接：https://arxiv.org/abs/2509.10919

【2】Rethinking Sparse Autoencoders: Select-and-Project for Fairness and Control from Encoder Features Alone
标题：重新思考稀疏自动编码器：选择和项目以实现公平性和仅通过编码器功能进行控制
链接：https://arxiv.org/abs/2509.10809

优化|敛散性(17篇)

【1】Foundational theory for optimal decision tree problems. II. Optimal hypersurface decision tree algorithm
标题：最优决策树问题的基础理论。二.最优超表面决策树算法
链接：https://arxiv.org/abs/2509.12057

【2】Low-rank Orthogonalization for Large-scale Matrix Optimization with Applications to Foundation Model Training
标题：大规模矩阵优化的低等级子化及其在基础模型训练中的应用
链接：https://arxiv.org/abs/2509.11983

【3】Topology Structure Optimization of Reservoirs Using GLMY Homology
标题：利用GLMY同调进行储层结构优化
链接：https://arxiv.org/abs/2509.11612

【4】Compressed Sensing: Mathematical Foundations, Implementation, and Advanced Optimization Techniques
标题：压缩感知：数学基础、实现和高级优化技术
链接：https://arxiv.org/abs/2509.11550

【5】Online Optimization on Hadamard Manifolds: Curvature Independent Regret Bounds on Horospherically Convex Objectives
标题：Hadamard流形上的在线优化：时球凸目标上曲率无关的遗憾界
链接：https://arxiv.org/abs/2509.11236

【6】Foundational theory for optimal decision tree problems. I. Algorithmic and geometric foundations
标题：最优决策树问题的基础理论。I.数学和几何基础
链接：https://arxiv.org/abs/2509.11226

【7】Harnessing Optimization Dynamics for Curvature-Informed Model Merging
标题：利用优化动力学进行曲线信息模型合并
链接：https://arxiv.org/abs/2509.11167

【8】DemandLens: Enhancing Forecast Accuracy Through Product-Specific Hyperparameter Optimization
标题：DemandLens：通过特定产品的超参数优化提高预测准确性
链接：https://arxiv.org/abs/2509.11085

【9】BERT4beam: Large AI Model Enabled Generalized Beamforming Optimization
标题：BERT 4beam：大型人工智能模型实现广义束形成优化
链接：https://arxiv.org/abs/2509.11056

【10】Optimal message passing for molecular prediction is simple, attentive and spatial
标题：分子预测的最佳信息传递是简单、专注且空间的
链接：https://arxiv.org/abs/2509.10871

【11】FACTORS: Factorial Approximation for Complementary Two-factor Optimization with Risk-aware Scoring
标题：因素：具有风险意识评分的互补双因素优化的因式逼近
链接：https://arxiv.org/abs/2509.10825

【12】Learning Concave Bid Shading Strategies in Online Auctions via Measure-valued Proximal Optimization
标题：通过测量值逼近优化学习在线拍卖中的凹凸投标着色策略
链接：https://arxiv.org/abs/2509.10693

【13】Optimal Multimarginal Schrödinger Bridge: Minimum Spanning Tree over Measure-valued Vertices
标题：最佳多边缘薛定汉桥：测量值点上的最小生成树
链接：https://arxiv.org/abs/2509.10626

【14】Designing MacPherson Suspension Architectures using Bayesian Optimization
标题：使用Bayesian优化设计麦克弗森悬架架构
链接：https://arxiv.org/abs/2206.09022

【15】Preconditioned subgradient method for composite optimization: overparameterization and fast convergence
标题：复合优化的预条件次梯度方法：过度参数化和快速收敛
链接：https://arxiv.org/abs/2509.11486

【16】From PowerSGD to PowerSGD+: Low-Rank Gradient Compression for Distributed Optimization with Convergence Guarantees
标题：从PowerSingapore到PowerSingapore+：具有收敛保证的分布式优化的低等级梯度压缩
链接：https://arxiv.org/abs/2509.11254

【17】Convergence Rate in Nonlinear Two-Time-Scale Stochastic Approximation with State (Time)-Dependence
标题：具有状态（时间）依赖性的非线性双时间尺度随机逼近的收敛速度
链接：https://arxiv.org/abs/2509.11039

预测|估计(13篇)

【1】Multimodal Regression for Enzyme Turnover Rates Prediction
标题：酶周转率预测的多峰回归
链接：https://arxiv.org/abs/2509.11782

【2】Beyond Regularity: Modeling Chaotic Mobility Patterns for Next Location Prediction
标题：超越规则：为下一个位置预测建模混乱移动模式
链接：https://arxiv.org/abs/2509.11713

【3】Know What You Don't Know: Selective Prediction for Early Exit DNNs
标题：知道你不知道的：提前退出DNN的选择性预测
链接：https://arxiv.org/abs/2509.11520

【4】Machine Learning-Driven Predictive Resource Management in Complex Science Workflows
标题：复杂科学工作流程中机器学习驱动的预测资源管理
链接：https://arxiv.org/abs/2509.11512

【5】BiLSTM-VHP: BiLSTM-Powered Network for Viral Host Prediction
标题：BiLSTM-VHP：用于病毒宿主预测的BiLSTM-VHP网络
链接：https://arxiv.org/abs/2509.11345

【6】Data-Efficient Ensemble Weather Forecasting with Diffusion Models
标题：使用扩散模型实现数据高效的综合天气预测
链接：https://arxiv.org/abs/2509.11047

【7】Hybrid Quantum Neural Networks for Efficient Protein-Ligand Binding Affinity Prediction
标题：用于高效蛋白质-配体结合亲和力预测的混合量子神经网络
链接：https://arxiv.org/abs/2509.11046

【8】California Wildfire Inventory (CAWFI): An Extensive Dataset for Predictive Techniques based on Artificial Intelligence
标题：加州野火清单（CALDI）：基于人工智能的预测技术的广泛数据集
链接：https://arxiv.org/abs/2509.11015

【9】GTS_Forecaster: a novel deep learning based geodetic time series forecasting toolbox with python
标题：GTS_Forecaster：一个新颖的基于深度学习的大地测量时间序列预测工具箱，使用Python
链接：https://arxiv.org/abs/2509.10560

【10】Multimodal Deep Learning for ATCO Command Lifecycle Modeling and Workload Prediction
标题：用于ATCO命令任务组建模和任务组预测的多模式深度学习
链接：https://arxiv.org/abs/2509.10522

【11】Gradient Estimation Methods of Approximate Multipliers for High-Accuracy Retraining of Deep Learning Models
标题：用于深度学习模型高准确性重新训练的逼近乘数的梯度估计方法
链接：https://arxiv.org/abs/2509.10519

【12】From Noise to Precision: A Diffusion-Driven Approach to Zero-Inflated Precipitation Prediction
标题：从噪声到精度：零膨胀降水预报的扩散驱动方法
链接：https://arxiv.org/abs/2509.10501

【13】Moment Estimates and DeepRitz Methods on Learning Diffusion Systems with Non-gradient Drifts
标题：具有非梯度漂移的学习扩散系统的矩估计和DeepRitz方法
链接：https://arxiv.org/abs/2509.10495

其他神经网络|深度学习|模型|建模(46篇)

【1】Learning Neural Networks by Neuron Pursuit
标题：通过Neuron Pursuit学习神经网络
链接：https://arxiv.org/abs/2509.12154

【2】Do machine learning climate models work in changing climate dynamics?
标题：机器学习气候模型在改变气候动态中是否有效？
链接：https://arxiv.org/abs/2509.12147

【3】A Time-Series Foundation Model by Universal Delay Embedding
标题：通用延迟嵌入的时间序列基础模型
链接：https://arxiv.org/abs/2509.12080

【4】Imitation Learning as Return Distribution Matching
标题：模仿学习作为回报分布匹配
链接：https://arxiv.org/abs/2509.12026

【5】Learning non-Markovian Dynamical Systems with Signature-based Encoders
标题：使用基于签名的编码器学习非马尔科夫动态系统
链接：https://arxiv.org/abs/2509.12022

【6】Learning from Uncertain Similarity and Unlabeled Data
标题：从不确定的相似性和未标记的数据中学习
链接：https://arxiv.org/abs/2509.11984

【7】Deep operator network for surrogate modeling of poroelasticity with random permeability fields
标题：基于深度算子网络的随机渗透率介质弹性模型
链接：https://arxiv.org/abs/2509.11966

【8】Data Fusion and Machine Learning for Ship Fuel Consumption Modelling - A Case of Bulk Carrier Vessel
标题：船舶油耗建模的数据融合和机器学习-以散货船为例
链接：https://arxiv.org/abs/2509.11750

【9】Analysing Python Machine Learning Notebooks with Moose
标题：用Moose分析Python机器学习笔记本
链接 ：https://arxiv.org/abs/2509.11748

【10】Fast and Interpretable Machine Learning Modelling of Atmospheric Molecular Clusters
标题：大气分子团簇的快速且可解释的机器学习建模
链接：https://arxiv.org/abs/2509.11728

【11】Disentangling Content from Style to Overcome Shortcut Learning: A Hybrid Generative-Discriminative Learning Framework
标题：将内容与风格区分开来以克服预设学习：混合生成歧视学习框架
链接：https://arxiv.org/abs/2509.11598

【12】Learning Singularity-Encoded Green's Functions with Application to Iterative Methods
标题：奇异编码格林函数的学习及其在迭代方法中的应用
链接：https://arxiv.org/abs/2509.11580

【13】CEMTM: Contextual Embedding-based Multimodal Topic Modeling
标题：CEMTM：基于上下文嵌入的多模式主题建模
链接：https://arxiv.org/abs/2509.11465

【14】Learning to Optimize Multi-Objective Alignment Through Dynamic Reward Weighting
标题：学习通过动态奖励加权优化多目标一致
链接：https://arxiv.org/abs/2509.11452

【15】Framing AI System Benchmarking as a Learning Task: FlexBench and the Open MLPerf Dataset
标题：将人工智能系统基准测试作为一项学习任务：FlexBench和开放MLPerf数据集
链接：https://arxiv.org/abs/2509.11413

【16】Enhancing ML Models Interpretability for Credit Scoring
标题：增强ML模型对信用评分的解释性
链接：https://arxiv.org/abs/2509.11389

【17】Efficient Single-Step Framework for Incremental Class Learning in Neural Networks
标题：神经网络中增量类学习的高效分步框架
链接：https://arxiv.org/abs/2509.11285

【18】PINGS: Physics-Informed Neural Network for Fast Generative Sampling
标题：PINGS：用于快速生成采样的物理信息神经网络
链接：https://arxiv.org/abs/2509.11284

【19】Revisiting Meter Tracking in Carnatic Music using Deep Learning Approaches
标题：使用深度学习方法重新审视狂欢音乐中的电表跟踪
链接：https://arxiv.org/abs/2509.11241

【20】RoVerFly: Robust and Versatile Learning-based Control of Quadrotor Across Payload Configurations
标题：RoVerFly：跨有效负载切换器的四螺旋桨鲁棒且多功能的基于学习的控制
链接：https://arxiv.org/abs/2509.11149

【21】An Advanced Convolutional Neural Network for Bearing Fault Diagnosis under Limited Data
标题：有限数据下轴承故障诊断的高级卷积神经网络
链接：https://arxiv.org/abs/2509.11053

【22】FragmentGPT: A Unified GPT Model for Fragment Growing, Linking, and Merging in Molecular Design
标题：FragmentGPT：分子设计中用于片段生长、连接和合并的统一GPT模型
链接：https://arxiv.org/abs/2509.11044

【23】Hardness, Structural Knowledge, and Opportunity: An Analytical Framework for Modular Performance Modeling
标题：硬度、结构知识和机会：模块化性能建模的分析框架
链接：https://arxiv.org/abs/2509.11000

【24】Decoupling Search and Learning in Neural Net Training
标题：神经网络训练中的搜索和学习脱钩
链接：https://arxiv.org/abs/2509.10973

【25】Clarifying Model Transparency: Interpretability versus Explainability in Deep Learning with MNIST and IMDB Examples
标题：澄清模型透明度：使用MNIST和IMDB示例深度学习中的可解释性与可解释性
链接：https://arxiv.org/abs/2509.10929

【26】RSL-RL: A Learning Library for Robotics Research
标题：RSL-RL：机器人研究的学习库
链接：https://arxiv.org/abs/2509.10771

【27】MinatoLoader: Accelerating Machine Learning Training Through Efficient Data Preprocessing
标题：MinatoPlayer：通过高效的数据预处理加速机器学习训练
链接：https://arxiv.org/abs/2509.10712

【28】DOSA: Differentiable Model-Based One-Loop Search for DNN Accelerators
标题：DOSA：基于差异模型的DNN加速器单循环搜索
链接：https://arxiv.org/abs/2509.10702

【29】Situation Model of the Transport, Transport Emissions and Meteorological Conditions
标题：交通、交通排放和气象条件的情景模型
链接：https://arxiv.org/abs/2509.10541

【30】Variational Gaussian Mixture Manifold Models for Client-Specific Federated Personalization
标题：用于客户特定联邦个性化的变分高斯混合Manifold模型
链接：https://arxiv.org/abs/2509.10521

【31】SOH-KLSTM: A Hybrid Kolmogorov-Arnold Network and LSTM Model for Enhanced Lithium-Ion Battery Health Monitoring
标题：SOH-KLSTM：用于增强锂离子电池健康状况监测的混合Kolmogorov-Arnold网络和LSTM模型
链接：https://arxiv.org/abs/2509.10496

【32】Agentic DDQN-Based Scheduling for Licensed and Unlicensed Band Allocation in Sidelink Networks
标题：侧链网络中基于DDQN的授权和非授权频段分配的统计调度
链接：https://arxiv.org/abs/2509.06775

【33】Information Entropy-Based Scheduling for Communication-Efficient Decentralized Learning
标题：基于信息熵的通信高效分散学习调度
链接：https://arxiv.org/abs/2507.17426

【34】Learning Stackable and Skippable LEGO Bricks for Efficient, Reconfigurable, and Variable-Resolution Diffusion Modeling
标题：学习可堆叠和可跳过的乐高积木，以实现高效、可重新配置和可变分辨率的扩散建模
链接：https://arxiv.org/abs/2310.06389

【35】The Morgan-Pitman Test of Equality of Variances and its Application to Machine Learning Model Evaluation and Selection
标题：方差相等的Morgan-Pitman检验及其在机器学习模型评估和选择中的应用
链接：https://arxiv.org/abs/2509.12185

【36】Quantum Noise Tomography with Physics-Informed Neural Networks
标题：使用物理信息神经网络的量子噪音断层扫描
链接：https://arxiv.org/abs/2509.11911

【37】SpaPool: Soft Partition Assignment Pooling for__Graph Neural Networks
标题：SpaPool：__图神经网络的软分区分配池
链接：https://arxiv.org/abs/2509.11675

【38】E-ROBOT: a dimension-free method for robust statistics and machine learning via Schrödinger bridge
标题：E-RObot：一种通过Schrödinger桥进行稳健统计和机器学习的无维方法
链接：https://arxiv.org/abs/2509.11532

【39】Quantum Architecture Search for Solving Quantum Machine Learning Tasks
标题：量子架构寻求解决量子机器学习任务
链接：https://arxiv.org/abs/2509.11198

【40】What is in a Price? Estimating Willingness-to-Pay with Bayesian Hierarchical Models
标题：什么是价格？利用Bayesian分层模型估计支付意愿
链接：https://arxiv.org/abs/2509.11089

【41】Kernel-based Stochastic Approximation Framework for Nonlinear Operator Learning
标题：基于核的非线性运算符学习随机逼近框架
链接：https://arxiv.org/abs/2509.11070

【42】Physics-informed neural network solves minimal surfaces in curved spacetime
标题：基于物理学的神经网络解决弯曲时空中的最小表面
链接：https://arxiv.org/abs/2509.10866

【43】On a Geometry of Interbrain Networks
标题：脑间网络的几何学
链接：https://arxiv.org/abs/2509.10650

【44】DeepSeasons: a Deep Learning scale-selecting approach to Seasonal Forecasts
标题：DeepSeasons：季节性预测的深度学习规模选择方法
链接：https://arxiv.org/abs/2509.10494

【45】Spectral Bottleneck in Deep Neural Networks: Noise is All You Need
标题：深度神经网络中的频谱瓶颈：噪音就是你所需要的一切
链接：https://arxiv.org/abs/2509.09719

【46】Green Learning for STAR-RIS mmWave Systems with Implicit CSI
标题：具有隐式SI的STAR-RIS毫米波系统的绿色学习
链接：https://arxiv.org/abs/2509.06820

其他(50篇)

【1】HoloGarment: 360° Novel View Synthesis of In-the-Wild Garments
标题：HoloGarment：野外服装的360°新颖视角合成
链接：https://arxiv.org/abs/2509.12187

【2】All that structure matches does not glitter
标题：所有匹配的结构都不会闪闪发光
链接：https://arxiv.org/abs/2509.12178

【3】Deceptive Risk Minimization: Out-of-Distribution Generalization by Deceiving Distribution Shift Detectors
标题：欺骗性风险最小化：通过欺骗分布漂移检测器实现分布外概括
链接：https://arxiv.org/abs/2509.12081

【4】Hi-DARTS: Hierarchical Dynamically Adapting Reinforcement Trading System
标题：Hi-DART：分层动态适应强化交易系统
链接：https://arxiv.org/abs/2509.12048

【5】Query-Focused Extractive Summarization for Sentiment Explanation
标题：以查询为中心的提取总结以进行情绪解释
链接：https://arxiv.org/abs/2509.11989

【6】Examining the Relationship between Scientific Publishing Activity and Hype-Driven Financial Bubbles: A Comparison of the Dot-Com and AI Eras
标题：审视科学出版活动与炒作驱动的金融泡沫之间的关系：Dot-Com和AI时代的比较
链接：https://arxiv.org/abs/2509.11982

【7】TabStruct: Measuring Structural Fidelity of Tabular Data
标题：TabStruct：测量表格数据的结构保真度
链接：https://arxiv.org/abs/2509.11950

【8】Transparent and Fair Profiling in Employment Services: Evidence from Switzerland
标题：就业服务透明公平的概况：来自瑞士的证据
链接：https://arxiv.org/abs/2509.11847

【9】Synthetic vs. Real Training Data for Visual Navigation
标题：视觉导航的合成与真实训练数据
链接：https://arxiv.org/abs/2509.11791

【10】User eXperience Perception Insights Dataset (UXPID): Synthetic User Feedback from Public Industrial Forums
标题：用户eXperience Percence Insights数据集（UXID）：来自公共工业论坛的合成用户反馈
链接：https://arxiv.org/abs/2509.11777

【11】Stabilizing PINNs: A regularization scheme for PINN training to avoid unstable fixed points of dynamical systems
标题：稳定PINN：PINN训练的正规化方案，以避免动态系统的不稳定不稳定点
链接：https://arxiv.org/abs/2509.11768

【12】Neural Audio Codecs for Prompt-Driven Universal Source Separation
标题：用于预算驱动通用源分离的神经音频编解码器
链接：https://arxiv.org/abs/2509.11717

【13】An Interventional Approach to Real-Time Disaster Assessment via Causal Attribution
标题：基于因果归因的实时灾害评估干预方法
链接：https://arxiv.org/abs/2509.11676

【14】Assessing On-the-Ground Disaster Impact Using Online Data Sources
标题：使用在线数据源评估实地灾害影响
链接：https://arxiv.org/abs/2509.11634

【15】Reasoned Safety Alignment: Ensuring Jailbreak Defense via Answer-Then-Check
标题：合理的安全调整：通过先检查确保越狱防御
链接：https://arxiv.org/abs/2509.11629

【16】AMLNet: A Knowledge-Based Multi-Agent Framework to Generate and Detect Realistic Money Laundering Transactions
标题：AMLNet：一个基于知识的多代理框架，用于生成和检测现实洗钱交易
链接：https://arxiv.org/abs/2509.11595

【17】RAPTOR: A Foundation Policy for Quadrotor Control
标题：RAPTOR：四螺旋桨控制的基础政策
链接：https://arxiv.org/abs/2509.11481

【18】Long-time dynamics and universality of nonconvex gradient descent
标题：非凸梯度下降的长期动力学和普适性
链接：https://arxiv.org/abs/2509.11426

【19】From Firewalls to Frontiers: AI Red-Teaming is a Domain-Specific Evolution of Cyber Red-Teaming
标题：从防火墙到前沿：人工智能红色团队是网络红色团队的特定领域进化
链接：https://arxiv.org/abs/2509.11398

【20】Decoding Musical Origins: Distinguishing Human and AI Composers
标题：解码音乐起源：区分人类和人工智能作曲家
链接：https://arxiv.org/abs/2509.11369

【21】Online Omniprediction with Long-Term Constraints
标题：具有长期约束的在线全方位预测
链接：https://arxiv.org/abs/2509.11357

【22】On Linear Mode Connectivity of Mixture-of-Experts Architectures
标题：专家混合架构的线性模式连通性
链接：https://arxiv.org/abs/2509.11348

【23】Opal: An Operator Algebra View of RLHF
标题：Opal：RL HF的运算子代数观点
链接：https://arxiv.org/abs/2509.11298

【24】SelectMix: Enhancing Label Noise Robustness through Targeted Sample Mixing
标题：SelectMix：通过有针对性的样本混合增强标签噪音稳健性
链接：https://arxiv.org/abs/2509.11265

【25】GK-SMOTE: A Hyperparameter-free Noise-Resilient Gaussian KDE-Based Oversampling Approach
标题：GK-SMOTE：一种无超参数、抗噪高斯ADE的过采样方法
链接：https://arxiv.org/abs/2509.11163

【26】Feature Space Topology Control via Hopkins Loss
标题：通过霍普金斯损失进行特征空间布局控制
链接：https://arxiv.org/abs/2509.11154

【27】Robustifying Diffusion-Denoised Smoothing Against Covariate Shift
标题：针对协变量漂移的Robustified扩散去噪平滑
链接：https://arxiv.org/abs/2509.10913

【28】Towards Automated Error Discovery: A Study in Conversational AI
标题：迈向自动错误发现：对话人工智能的研究
链接：https://arxiv.org/abs/2509.10833

【29】Contextual Budget Bandit for Food Rescue Volunteer Engagement
标题：食品救援志愿者参与的背景预算强盗
链接：https://arxiv.org/abs/2509.10777

【30】Pluralistic Alignment for Healthcare: A Role-Driven Framework
标题：医疗保健多元化协调：角色驱动框架
链接：https://arxiv.org/abs/2509.10685

【31】pySigLib - Fast Signature-Based Computations on CPU and GPU
标题：pySigLib -在中央处理器和图形处理器上进行基于签名的快速计算
链接：https://arxiv.org/abs/2509.10613

【32】National Running Club Database: Assessing Collegiate Club Athletes' Cross Country Race Results
标题：国家跑步俱乐部数据库：评估大学俱乐部运动员的越野赛成绩
链接：https://arxiv.org/abs/2509.10600

【33】Auditable Early Stopping for Agentic Routing: Ledger-Verified Run-Wise Certificates under Local DP
标题：可审核的公共路由早期停止：本地DP下的Ledger验证运行明智证书
链接：https://arxiv.org/abs/2509.10550

【34】Contextuality, Holonomy and Discrete Fiber Bundles in Group-Valued Boltzmann Machines
标题：群值Boltzmann机中的上下文、完整性和离散纤维束
链接：https://arxiv.org/abs/2509.10536

【35】Decoupling the "What" and "Where" With Polar Coordinate Positional Embeddings
链接：https://arxiv.org/abs/2509.10534

【36】Mitigating Catastrophic Forgetting and Mode Collapse in Text-to-Image Diffusion via Latent Replay
标题：通过潜在重播减轻文本到图像扩散中的灾难性遗忘和模式崩溃
链接：https://arxiv.org/abs/2509.10529

【37】Mixture-of-Clustered-Experts: Advancing Expert Specialization and Generalization in Instruction Tuning
标题：混合型专家：推进指令调优中的专家专业化和通用化
链接：https://arxiv.org/abs/2509.10513

【38】AttnBoost: Retail Supply Chain Sales Insights via Gradient Boosting Perspective
标题：AttnBoot：通过梯度提升视角洞察零售供应链销售
链接：https://arxiv.org/abs/2509.10506

【39】Identifiable Autoregressive Variational Autoencoders for Nonlinear and Nonstationary Spatio-Temporal Blind Source Separation
标题：用于非线性和非平稳时空盲源分离的可识别自回归变分自动编码器
链接：https://arxiv.org/abs/2509.11962

【40】ProteuS: A Generative Approach for Simulating Concept Drift in Financial Markets
标题：ProteuS：模拟金融市场概念漂移的生成方法
链接：https://arxiv.org/abs/2509.11844

【41】A Particle-Flow Algorithm for Free-Support Wasserstein Barycenters
标题：自由支持Wasserstein重心的粒子流算法
链接：https://arxiv.org/abs/2509.11435

【42】Some Robustness Properties of Label Cleaning
标题：标签清洁的一些鲁棒性
链接：https://arxiv.org/abs/2509.11379

【43】Investigating the Lottery Ticket Hypothesis for Variational Quantum Circuits
标题：研究变分量子电路的彩票假说
链接：https://arxiv.org/abs/2509.11190

【44】Maximum diversity, weighting and invariants of time series
标题：时间序列的最大多样性、加权和不变量
链接：https://arxiv.org/abs/2509.11146

【45】Gradient Methods with Online Scaling Part II. Practical Aspects
标题：具有在线缩放的梯度方法第二部分。实际方面
链接：https://arxiv.org/abs/2509.11007

【46】Variable Selection Using Relative Importance Rankings
标题：使用相对重要性排名选择变量
链接：https://arxiv.org/abs/2509.10853

【47】Trial-Level Time-frequency EEG Desynchronization as a Neural Marker of Pain
标题：试验级时频脑电去同步化作为疼痛的神经标记物
链接：https://arxiv.org/abs/2509.10552

【48】Biomarkers of brain diseases
标题：脑部疾病的生物标志物
链接：https://arxiv.org/abs/2509.10547

【49】An Interpretable Ensemble Framework for Multi-Omics Dementia Biomarker Discovery Under HDLSS Conditions
标题：HDLSS条件下多组性痴呆生物标志物发现的可解释集合框架
链接：https://arxiv.org/abs/2509.10527

【50】YOLO-based Bearing Fault Diagnosis With Continuous Wavelet Transform
标题：基于YOLO的连续小波变换轴承故障诊断
链接：https://arxiv.org/abs/2509.03070

机器翻译由腾讯交互翻译提供，仅供参考

点击“阅读原文”获取带摘要的学术速递