2026-07论文中文摘要
AuthProbe：基于规范驱动的多身份检测招聘 API 中对象级授权漏洞
AuthProbe: Specification-Driven, Multi-Identity Detection of Broken Object-Level Authorization in Recruitment API
arXiv 2607.20574 · 2026-07-24单细胞基础模型基准中的预训练污染审计
Auditing pretraining contamination in single-cell foundation model benchmarks
arXiv 2607.20572 · 2026-07-24关于格点集和集大小的一个问题
A problem on sumset sizes of sets of lattice points
arXiv 2607.20571 · 2026-07-24线性磁电体和等离子体中光传播的引力效应
Gravitational Effects on Light Propagation in Linear Magnetoelectrics and Plasmas
arXiv 2607.20568 · 2026-07-24遥平行高斯-博内特膜中的张量共振
Tensor resonances in teleparallel Gauss-Bonnet branes
arXiv 2607.20567 · 2026-07-24五阶彼得森击中问题、第五个辛格转移以及无定向配边中的几何生成元
The rank-five Peterson hit problem, the fifth Singer transfer, and a geometric generator in unoriented cobordism
arXiv 2607.20566 · 2026-07-24纽曼 - 贾尼斯变形的阻碍与剩余完备理论
An obstruction and residual-completion theory for Newman--Janis deformations
arXiv 2607.20565 · 2026-07-24向量空间的参数理论及其在模糊向量空间中的应用
A Parametric Theory of Vector Spaces with Applications to Fuzzy Vector Spaces
arXiv 2607.20563 · 2026-07-24关于缩减索姆博尔指数的精确界
Sharp Bounds on Diminished Sombor Index
arXiv 2607.20562 · 2026-07-24Chronofy：一种用于时间感知检索增强生成中信息有效性的时态逻辑衰减架构
Chronofy: A Temporal-Logical Decay Architecture for Information Validity in Time-Aware Retrieval-Augmented Generation
arXiv 2607.20560 · 2026-07-24印度社会经济指标自动编码器辅助降尺度中地理空间和人口普查代理的联合利用（JUGAAD）
Joint Utilization of Geospatial and census proxies for Autoencoder-Assisted Downscaling (JUGAAD) of socioeconomic indicators in India
arXiv 2607.20559 · 2026-07-24StabilityBench：评估大语言模型中的不稳定性
StabilityBench: Benchmarking Instability in LLMs
arXiv 2607.20558 · 2026-07-24孙悟空爆炸：一个统一的科学多模态基础模型
Monkey King Bang: A Unified Scientific Multimodal Foundation Model
arXiv 2607.20557 · 2026-07-24双评分：可靠提取强彩票票号
Double-Scoring: Reliable Extraction of Strong Lottery Tickets
arXiv 2607.20555 · 2026-07-24基于学习优化图神经网络的智能城市NR-V2X网络的人工智能驱动多跳中继选择
AI-Driven Multi-Hop Relay Selection for Smart Urban NR-V2X Networks via Learning-to-Optimize Graph Neural Networks
arXiv 2607.20554 · 2026-07-24热力学权重衰减：通过注意力比热探索快速学习加速
Thermodynamic Weight Decay: Exploring Grokking Acceleration via Attention Specific Heat
arXiv 2607.20552 · 2026-07-24SenCos-GEM：用于属性预测的SENet校准和余弦定律约束的几何增强分子表示
SenCos-GEM: SENet-Calibrated and Law-of-Cosines-Constrained Geometry-Enhanced Molecular Representation for Property Prediction
arXiv 2607.20551 · 2026-07-24超越基于结构的药物设计：多药理学和多靶点药物设计中的几何深度学习
Beyond SBDD: Geometric Deep Learning in Polypharmacology and Multi-target Drug Design
arXiv 2607.20550 · 2026-07-24SevDiff：用于长尾冲突轨迹生成的严重程度条件扩散
SevDiff: Severity-Conditioned Diffusion for Long-Tail Conflict Trajectory Generation
arXiv 2607.20549 · 2026-07-24SOAP、Muon及其他：推动语言模型预训练规模
SOAP, Muon, and Beyond: Pushing LLM Pretraining Scales
arXiv 2607.20548 · 2026-07-24使用多智能体强化学习解决空中走廊降级监视下的冲突
Conflict Resolution under Degraded Surveillance in Air Corridors Using Multi-Agent Reinforcement Learning
arXiv 2607.20547 · 2026-07-24振荡磁场对基于隐花色素磁受体中自由基对电偶极矩的影响
Influence of Oscillating Magnetic Fields on the Electric Dipole Moment of Radical Pairs in Cryptochrome Based Magnetoreception
arXiv 2607.20546 · 2026-07-24StrideDiffusion：加速用于时间序列生成的扩散模型
StrideDiffusion: Accelerating Diffusion Models for Time-series Generation
arXiv 2607.20545 · 2026-07-24当可验证奖励的强化学习缩小推理边界时：诊断Pass@k反转
When RLVR Shrinks the Reasoning Boundary: Diagnosing Pass@k Inversion
arXiv 2607.20543 · 2026-07-24通过决策感知机器学习改善基本药物的获取
Improving Access to Essential Medicines via Decision-Aware Machine Learning
arXiv 2607.20542 · 2026-07-24HypNO：一种基于图的神经算子，用于双曲守恒律的物理信息消息传递
HypNO: A Graph-Based Neural Operator with Physics-Informed Message Passing for Hyperbolic Conservation Laws
arXiv 2607.20541 · 2026-07-24从原子到熵：凸域中扩散训练的最优噪声分配
From Atoms to Entropy: Optimal Noise Allocation for Diffusion Training in the Convex Regime
arXiv 2607.20540 · 2026-07-24利用生物动力学知识先验进行数据稀缺的生物过程建模
Leveraging Biokinetic Knowledge Priors for Data-Scarce Bioprocess Modeling
arXiv 2607.20539 · 2026-07-24编解码器规范：为Transformer KV缓存学习压缩友好型规范
Codec-Gauge: Learning Compression-Friendly Gauges for Transformer KV Caches
arXiv 2607.20538 · 2026-07-24可靠表格问答：可靠性标注需要多少监督？
ReliableTableQA:How Much Supervision Does Reliability Annotation Need?
arXiv 2607.20537 · 2026-07-24AppWorld-UL：用于工具使用的多样化智能体-用户交互基准测试
AppWorld-UL: Benchmarking Diverse Agent-User Interactions for Tool-Use
arXiv 2607.20536 · 2026-07-24一种用于零样本数字孪生的图神经网络方法
A Graph Neural Network approach to zero-shot Digital Twins
arXiv 2607.20535 · 2026-07-24量子纠错与生物纠错：量子比特与神经元之间的结构类比
Quantum error correction and biological error correction: A structural analogy between qubits and neurons
arXiv 2607.20534 · 2026-07-24在布莱克-利特曼模型中锚定投资者观点：神经谓词
Grounding Investor Views: Neural Predicates in the Black-Litterman Model
arXiv 2607.20533 · 2026-07-24立场：停止每次都被动地修补模型，开始主动的测试驱动人工智能开发
Position: Stop Reactively Patching Your Model Every Time and Start Proactive Test-Driven AI Development
arXiv 2607.20532 · 2026-07-24CLOE：用于异常检测的克里斯托费尔损失自动编码器
CLOE: Christoffel Loss Autoencoder for Anomaly Detection
arXiv 2607.20530 · 2026-07-24通过结构化专家判断对多语言模型系统进行不确定性感知信任估计
Uncertainty-Aware Trust Estimation for Multi-LLM Systems via Structured Expert Judgement
arXiv 2607.20529 · 2026-07-24PromptPack：扩展用于在线推荐的大语言模型注释代理
PromptPack: Scaling LLM Annotation Agents for Online Recommendation
arXiv 2607.20528 · 2026-07-24评估和保障智能科学合成中的引用忠实性
Evaluating and Guarding Citation Faithfulness in Agentic Scientific Synthesis
arXiv 2607.20527 · 2026-07-24ConfidenceBench：评估大语言模型中的置信度校准
ConfidenceBench: Evaluating Confidence Calibration in Large Language Models
arXiv 2607.20526 · 2026-07-24注意力退化、功能令牌锚定以及基于注意力的大语言模型干预的局限性
Attention Degradation, Function Token Anchoring, and the Limits of Attention-Based Intervention in Large Language Models
arXiv 2607.20524 · 2026-07-24主动SAE特征平面是否具有更多的全同性？Gemma中的预注册反转
Do Active SAE Feature Planes Carry More Holonomy? A Preregistered Reversal in Gemma
arXiv 2607.20522 · 2026-07-24用于状态估计的生成式贝叶斯滤波
Generative Bayesian Filtering for State Estimation
arXiv 2607.20521 · 2026-07-24数学问题解决中大型语言模型在可执行推理约束下的表示鲁棒性
Representation Robustness Under Executable Reasoning Constraints in Large Language Models for Mathematical Problem Solving
arXiv 2607.20520 · 2026-07-24循环变换器中的自适应深度：诊断学习到的停止门和轨迹读出
Adaptive Depth in Looped Transformers: Diagnosing Learned Halting Gates and Trajectory Readouts
arXiv 2607.20519 · 2026-07-24CANN基准测试：针对真实NPU和算法限制对代理生成的内核进行基准测试
CANN Bench: Benchmarking Agent Generated Kernels against Real NPU and Algorithmic Limits
arXiv 2607.20518 · 2026-07-24多模态CoLRAG-TF：复杂PDF的三重过滤检索
Multimodal CoLRAG-TF: Triple-Filtered Retrieval for Complex PDFs
arXiv 2607.20517 · 2026-07-24使用大语言模型扩展闭环特征通道配置
Scaling Closed-Loop Feature Channel Configuration with LLMs
arXiv 2607.20516 · 2026-07-24从不一致的人类反馈中实现可靠性感知的大语言模型对齐
Reliability-Aware LLM Alignment from Inconsistent Human Feedback
arXiv 2607.20515 · 2026-07-24基于张拉整体翼结构的鳐鱼类运动动力学建模与分析
Dynamics modeling and analysis of batoid-type locomotion powered by tensegrity wing structure
arXiv 2607.20514 · 2026-07-24
