2026-07论文中文摘要
AINTMA：用于自主测试管理的智能AI架构，具备生成式智能、安全云通信和自适应质量分析
AINTMA: Agentic AI Architecture for Autonomous Test Management with Generative Intelligence, Secure Cloud Communication and Adaptive Quality Analytics
arXiv 2607.20452 · 2026-07-24语义场理论：历史起源、高阶交互与稳定语义推理
Semantic Field Theory: Historical Origin, Higher-Order Interaction, and Stabilized Semantic Inference
arXiv 2607.20451 · 2026-07-24ShriNep@EEUCA 2026：RAKSHAK - 用于毒性意图分类的具有原理蒸馏和拼图增强训练的多任务DeBERTa
ShriNep@EEUCA 2026: RAKSHAK - Multi-Task DeBERTa with Rationale Distillation and Jigsaw-Augmented Training for Toxic Intent Classification
arXiv 2607.20450 · 2026-07-24模型中的讲述者：大语言模型中的叙事模式继承、升级动态和对齐治理
The Storyteller in the Model: Narrative Pattern Inheritance, Escalation Dynamics, and Alignment Governance in LLMs
arXiv 2607.20449 · 2026-07-24Domyn-Small：一个欧洲的100亿参数推理语言模型
Domyn-Small: A European 10B Reasoning Language Model
arXiv 2607.20448 · 2026-07-24thaulab@EEUCA 2026：谁对谁说了什么？一种用于游戏毒性检测的目标感知神经符号管道
thaulab@EEUCA 2026: Who Said What to Whom? A Targeting-Aware Neural-Symbolic Pipeline for Gaming Toxicity Detection
arXiv 2607.20447 · 2026-07-24区分人工与真实：评估大语言模型检测由大语言模型生成的内容
Distinguishing Artificial from Authentic: Evaluating LLMs for Detecting LLM-Generated Content
arXiv 2607.20446 · 2026-07-24SCoPE：用于对话中情感识别的转移感知说话者条件先验
SCoPE: Shift-Aware Speaker-Conditioned Priors for Emotion Recognition in Conversations
arXiv 2607.20445 · 2026-07-24GLAN-QnA-KR：一个无种子分类法驱动的韩语指令语料库
GLAN-QnA-KR: A Seedless Taxonomy-Driven Korean Instruction Corpus
arXiv 2607.20443 · 2026-07-24Naver-News-KO：用于摘要模型开源微调的韩语新闻摘要数据集
Naver-News-KO: A Korean News Summarization Dataset for Open-Source Fine-Tuning of Summarization Models
arXiv 2607.20442 · 2026-07-24大语言模型世界模型中的信念传播：用预测市场测量战略信息偏差
Belief Propagation in LLM World Models: Measuring Strategic Information Bias with Prediction Markets
arXiv 2607.20441 · 2026-07-24先回答再编辑：用于保留效用的反蒸馏的推理骨架编辑
Answer-then-Edit: Reasoning Skeleton Editing for Anti-Distillation with Preserved Utility
arXiv 2607.20440 · 2026-07-24SemEval-2026任务6中的AsymVerify：用于政治逃避检测的非对称置信门控验证
AsymVerify at SemEval-2026 Task 6: Asymmetric Confidence-Gated Verification for Political Evasion Detection
arXiv 2607.20439 · 2026-07-24作为频谱更新重组的偏好调整
Preference Tuning as Spectral Update Reorganization
arXiv 2607.20438 · 2026-07-24TopoGuard：基于图论的针对RAG分裂知识攻击的防御方法
TopoGuard: Graph Theory Based Defenses Against Split-Knowledge Attacks on RAG
arXiv 2607.20437 · 2026-07-24路由子空间：审计微调语言模型中评估到部署的不匹配
Routing Subspaces: Auditing Evaluation-to-Deployment Mismatch in Fine-Tuned Language Models
arXiv 2607.20436 · 2026-07-24使开源文本语言模型水印在合并后仍具耐久性
Making Open-Source Text LLM Watermarks Durable Against Merging
arXiv 2607.20435 · 2026-07-24突破压缩瓶颈：从理论到实践
Break Through the Compression Bottleneck: From Theory to Practice
arXiv 2607.20434 · 2026-07-24莫尔：让模型为稳健的跨域知识编辑指引自身方向
Moir: Let the Model Direct Its Own Story for Robust Cross-Domain Knowledge Editing
arXiv 2607.20433 · 2026-07-24立场：自然语言不应完全取代形式语言
Position: Natural Language Should Not Fully Replace Formal Languages
arXiv 2607.20432 · 2026-07-24用于证据感知材料文献分析的技能收缩代理
Skill-Contracted Agents for Evidence-Aware Materials Literature Analysis
arXiv 2607.20431 · 2026-07-24LLM-INSTRUCT在UZH 2026共享任务中的应用：用于段落级论证挖掘的约束感知检索和选择性辩论
LLM-INSTRUCT at UZH Shared Task 2026: Constraint-Aware Retrieval and Selective Debate for Paragraph-Level Argument Mining
arXiv 2607.20430 · 2026-07-24更多并非更好：大语言模型观点多样性的关键因素是什么？
More Is Not More: What Matters for Diversity in LLM Opinions?
arXiv 2607.20429 · 2026-07-24用于识别皮肤免疫相关不良事件的人在回路大语言模型框架
Human-in-the-Loop Large Language Model Framework for Identification of Cutaneous Immune-Related Adverse Events
arXiv 2607.20428 · 2026-07-24混合专家路由是哈夫曼编码吗？发现思维链中的频率多样性定律
Is MoE Routing a Huffman Code? Discovering the Frequency-Diversity Law in Chain-of-Thought
arXiv 2607.20427 · 2026-07-24知识注入存在于混合专家模型中吗？探索混合专家模型中基于专家感知的对比解码以减轻大语言模型的幻觉
Knowledge Injection Exists in MoE? Exploring Expert-Aware Contrast Decoding in MoE for Mitigating LLMs'Hallucinations
arXiv 2607.20426 · 2026-07-24什么是好的？从大语言模型推理痕迹中提取并测试文学质量的隐含理论
What is Good? Extracting and Testing Implicit Theories of Literary Quality from LLM Reasoning Traces
arXiv 2607.20425 · 2026-07-24CYGNO探测器对暗物质的灵敏度：采用HFO-1234ze增强气体混合物
Dark Matter Sensitivity of the CYGNO Detector with HFO-1234ze Enhanced Gas Mixtures
arXiv 2607.20370 · 2026-07-24通过签名诱导最优传输的模型风险
Path-Space Model Risk via Signature-Induced Optimal Transport
arXiv 2607.20343 · 2026-07-24用于低温探测器校准的低温可切换电子源的演示
Demonstration of a cryogenic, switchable electron source for low-temperature detector calibration
arXiv 2607.20324 · 2026-07-24用于耦合热流体场的注意力图神经网络的无标签有限体积残差训练
Label-Free Finite-Volume-Residual Training of Attention Graph Neural Networks for Coupled Thermo-Fluid Fields
arXiv 2607.20321 · 2026-07-24O(3)无尺度引力中的暗能量与暗物质相互作用
Interacting Dark Energy and Dark Matter in O(3) No-Scale Gravity
arXiv 2607.20320 · 2026-07-24弹性块存储的黑盒性能评估：契约、速率限制模型与软件探索
Black-Box Performance Evaluation of Elastic Block Storage: Contract, Rate-Limiting Model, and Software Exploration
arXiv 2607.20319 · 2026-07-24Eos探测器：混合光学探测技术的演示器
The Eos detector: a demonstrator of hybrid optical detection technology
arXiv 2607.20285 · 2026-07-24基于实例硬度的不平衡回归相关性
Instance Hardness-Based Relevance for Imbalanced Regression
arXiv 2607.20173 · 2026-07-24跨项目缺陷预测的多阶段动态选择
Multi-stage Dynamic Selection for Cross-Project Defect Prediction
arXiv 2607.20151 · 2026-07-24开放技能风险：使用现实世界中的危险第三方技能时对智能体安全性进行基准测试
OpenSkillRisk: Benchmarking Agent Safety When Using Real-World Risky Third-Party Skills
arXiv 2607.20121 · 2026-07-24通过实验室实验对运行中的波浪诱导冰侵蚀模型进行理论开发
Theoretical development of an operational wave-induced ice erosion model through laboratory experiments
arXiv 2607.20118 · 2026-07-24通过动态评分标准共同进化大语言模型评估器和策略
Co-Evolving LLM Evaluators and Policies via DynamicRubric
arXiv 2607.20083 · 2026-07-24PRO-LONG：程序化内存实现长程推理
PRO-LONG: Programmatic Memory Enables Long-Horizon Reasoning
arXiv 2607.20064 · 2026-07-24\(\mathbb{C}^5\)中雅可比映射全局单射性的一个反例及其解析根
A Counterexample to the Global Injectivity of a Jacobian Mapping in \mathbb{C}^5 and Its Analytical Roots
arXiv 2607.20049 · 2026-07-24基于广义卡尔曼滤波器的时间差分强化学习
Generalized Kalman filter based temporal difference reinforcement learning
arXiv 2607.20010 · 2026-07-24迈向不同区域类型中稳健的深度学习哨兵2建筑物检测的季节性指南
Toward Seasonal Guidelines for Robust Deep-Learning Sentinel-2 Building Detection in Different Area Types
arXiv 2607.19994 · 2026-07-24利用非冗余孔径干涉测量作为同步加速器光表征的诊断工具
Exploiting Non Redundant Aperture Interferometry as a Diagnostics Tool for Synchrotron Light Characterization
arXiv 2607.19991 · 2026-07-24UniRank：用于统一序列建模和特征交互的排序模型基准测试
UniRank: Benchmarking Ranking Models for Unified Sequential Modeling and Feature Interaction
arXiv 2607.19987 · 2026-07-24用于束流尺寸测量的非冗余孔径干涉测量法的能力与局限
Capabilities and Limitations of Non-Redundant Aperture Interferometry for Beam Size Measurements
arXiv 2607.19976 · 2026-07-24关于量子拉丁方的可能基数
On the possible cardinalities of quantum Latin squares
arXiv 2607.19969 · 2026-07-24局部有限三角范畴中的高阶簇倾斜对象
Higher cluster tilting objects in locally finite triangulated categories
arXiv 2607.19916 · 2026-07-24具有多个不可穿透障碍物的多域有限元-边界元耦合
Multi-domain FEM-BEM coupling with several impenetrable obstacles
arXiv 2607.19896 · 2026-07-24OPIUM：通过双目标潜在优化减轻引导外部性和过度拒绝
OPIUM: Mitigating Steering Externalities and Over-Refusal via Dual Objective Latent Optimization
arXiv 2607.19806 · 2026-07-24
