2026-07论文中文摘要
评估残差：长时序评估在匹配短任务性能之外的补充作用
Benchmarking the Residual: What Long-Horizon Evaluations Add Beyond Matched Short-Task Performance
arXiv 2607.27283 · 2026-07-31训练动力学：语言模型中涌现、可塑性丧失与回路控制的成核速率定律
The Kinetics of Training: A Driven-Nucleation Rate Law for Emergence, Plasticity Loss, and Circuit Control in Language Models
arXiv 2607.27281 · 2026-07-31各向同性高斯动态下鲁棒随时间优化的期望生存时间界
Expected Survival-Time Bounds for Robust Optimization Over Time under Isotropic Gaussian Dynamics
arXiv 2607.27280 · 2026-07-31环的导数何时容许指数映射？
When does a derivation of a ring admit the exponential?
arXiv 2607.27279 · 2026-07-31可加可分享乐博弈中强受欢迎性的复杂性
Complexity of Strong Popularity in Additively Separable Hedonic Games
arXiv 2607.27277 · 2026-07-31利用ATLAS探测器在质心系能量为13 TeV和13.6 TeV的质子-质子碰撞中，对H→ττ衰变道下的t\bar{t}H与tH产生过程的研究
Study of $t\bar{t}H$ and $tH$ production in the $H\toττ$ channel in $pp$ collisions at $\sqrt{s}=13$ TeV and 13.6 TeV with the ATLAS detector
arXiv 2607.27276 · 2026-07-31Flat Score, Amplified Failures：误差预算如何掩盖量化大语言模型智能体中的损害
Flat Score, Amplified Failures: How the Error Budget Masks Damage in Quantized LLM Agents
arXiv 2607.27275 · 2026-07-31SDO：面向高效大语言模型后训练的结构感知数据组织
SDO: Structure-Aware Data Organization for Efficient LLM Post-Training
arXiv 2607.27273 · 2026-07-31多圆柱上多重调和映射的系数估计与Bohr现象
Coefficient estimates and Bohr phenomenon for pluriharmonic mappings in the polydisc
arXiv 2607.27272 · 2026-07-31RLPF：基于性能反馈的代码生成强化学习
RLPF: Reinforcement Learning from Performance Feedback for Code Generation
arXiv 2607.27271 · 2026-07-31BMOA：编译器诱导数值偏差的基线-机制-结果归因
BMOA: Baseline-Mechanism-Outcome Attribution for Compiler-Induced Numerical Deviations
arXiv 2607.27270 · 2026-07-31超越KV重建：推测解码中MLA草稿模型的功能重建
Beyond KV Reconstruction: Functional Reconstruction for MLA Draft Models in Speculative Decoding
arXiv 2607.27269 · 2026-07-31FAVA：基于证据支持的权限图的经验证智能体的形式化授权
FAVA: Formal Authorization for Verified Agents with Evidence-Backed Permission Graphs
arXiv 2607.27267 · 2026-07-31大规模戏剧小册子：排版的统计比较分析
Theatre Chapbooks At Scale: A Statistical Comparative Analysis of Typography
arXiv 2607.27266 · 2026-07-31PlatformBid：来自统一广告平台视角的自动竞价基准
PlatformBid: An Auto-Bidding Benchmark from a Unified Advertising Platform's Perspective
arXiv 2607.27265 · 2026-07-31利用循环对称性得到树覆盖的更强下界
Stronger Lower Bounds for Tree Covers via Cyclic Symmetry
arXiv 2607.27264 · 2026-07-31DoTime：用于干预和反事实时间序列的合成基准生成器
DoTime: A Synthetic Benchmark Generator for Interventional and Counterfactual Time Series
arXiv 2607.27263 · 2026-07-31大语言模型引导的初始化用于加速混合量子-经典医学图像分类
LLM-Guided Initialization for Accelerated Hybrid Quantum-Classical Medical Image Classification
arXiv 2607.27262 · 2026-07-31不仅仅是更多演示：用于数据高效的鲁棒机器人模仿的反事实动作敏感性覆盖
It's Not Just More Demos: Counterfactual Action Sensitivity Coverage for Data-Efficient Robust Robot Imitation
arXiv 2607.27261 · 2026-07-31CircuitProver：基于可复用电路证明库的智能体Lean 4定理证明框架，用于硬件验证
CircuitProver: Agentic Lean 4 Theorem Proving with Reusable Circuit Proof Library for Hardware Verification
arXiv 2607.27259 · 2026-07-31PlantBGC：基于无标签领域自适应与弱监督的植物生物合成基因簇发现Transformer
PlantBGC: Transformer for Plant BGC Discovery via Label-Free Domain Adaptation and Weak Supervision
arXiv 2607.27258 · 2026-07-31怪兽群是戈德施密特G3融合的一个完备化
The Monster group is a completion of the Goldschmidt G3-amalgam
arXiv 2607.27256 · 2026-07-31更多数据会导致更差的决策？Gram不相容下神经网络的偏好反转
More Data, Worse Decisions? Preference Reversals in Neural Networks under Gram Incompatibility
arXiv 2607.27255 · 2026-07-31地基引力波探测器网络的协同设计
Co-design of ground-based gravitational wave detector networks
arXiv 2607.27254 · 2026-07-31物质框架：来自慢力载体的硬反冲
A Material Frame: Hard Recoils from Slow Force Carriers
arXiv 2607.27253 · 2026-07-31双边逻辑中强否定的可定义性
On the Definability of Strong Negation in Bilateral Logics
arXiv 2607.27252 · 2026-07-31用于半导体热机械可靠性的递归Transformer
Recursive transformers for semiconductor thermo-mechanical reliability
arXiv 2607.27251 · 2026-07-31上下文文件对编码智能体有帮助吗？针对真实仓库的双智能体消融研究
Do Context Files Help Coding Agents? A Two-Agent Ablation Study on Real Repositories
arXiv 2607.27250 · 2026-07-31OwlPath：面向大语言模型漏洞修复的无损知识压缩
OwlPath: Lossless Knowledge Compression for LLM Bug Repair
arXiv 2607.27249 · 2026-07-31分歧解码：无训练的能力融合
Divergence Decoding: Training-Free Capability Fusion
arXiv 2607.27248 · 2026-07-31为量子计算设计紧框架
Designing tight frames for quantum computing
arXiv 2607.27247 · 2026-07-312-控制数与上中位度：Graffiti.pc 猜想387的证明
The 2-Domination Number and the Upper Median Degree: A Proof of Graffiti.pc Conjecture 387
arXiv 2607.27246 · 2026-07-31小型复合系统的坐标与动量分布
Coordinate and Momentum Distributions of a Small Composite System
arXiv 2607.27244 · 2026-07-31带非线性摩擦的机电振荡器中的多分支参数共振与燕尾突变
Multibranched parametric resonance and swallowtail catastrophe in electromechanical oscillators with nonlinear friction
arXiv 2607.27242 · 2026-07-31关于余割幂的拉马努金型主定理猜想的证明
Proof of a Conjectured Ramanujan Type Master Theorem for Powers of the Cosecant
arXiv 2607.27241 · 2026-07-31模型合并中的非对称坍塌：当拒绝覆盖识别
Asymmetric Collapse in Model Merging: When Refusal Over- writes Recognition
arXiv 2607.27240 · 2026-07-31参考依赖与WTA/WTP差距的结构
Reference Dependence and the Structure of the WTA/WTP Gap
arXiv 2607.27239 · 2026-07-31依赖参照的效用理论
A Theory of Reference-Dependent Utility
arXiv 2607.27238 · 2026-07-31随机动力系统的几何方法
Geometric Methods for Stochastic Dynamical Systems
arXiv 2607.27237 · 2026-07-31数据场论：用于信号检测的泛函重整化群理论与应用
Data Field Theory: Theory and Applications of the Functional Renormalization Group for Signal Detection
arXiv 2607.27236 · 2026-07-31RadHarmony：智能体AI时代的放射数据处理方案
RadHarmony: Radiological Data Handling in the Era of Agentic AI
arXiv 2607.27235 · 2026-07-31LoKI-GM：用于等离子体化学研究的全局模型框架
LoKI-GM: a global model framework for plasma chemistry studies
arXiv 2607.27234 · 2026-07-31AI辅助的开源软件提交预评审：来自BOSC 2026的经验报告
AI-assisted pre-review of open-source software submissions: an experience report from BOSC 2026
arXiv 2607.27228 · 2026-07-31除环上带常数项的多项式映射与广义Kaplansky-L'vov猜想
Polynomial Maps with Constants over Division Algebras and the Generalized Kaplansky--L'vov Conjecture
arXiv 2607.27226 · 2026-07-31矩阵环上的某些函数恒等式
Certain functional identities on matrix rings
arXiv 2607.27225 · 2026-07-31Psych-ECA：用于纵向精神病学中合成对照组的可复现半合成基准
Psych-ECA: A Reproducible Semi-Synthetic Benchmark for Synthetic Control Arms in Longitudinal Psychiatry
arXiv 2607.27224 · 2026-07-31关于phi-(k,n)-吸收δ-准素超理想
On phi-(k,n)-absorbing delta-primary hyperideals
arXiv 2607.27223 · 2026-07-31度量空间中的线段与凸性
Segments and Convexity in Metric Spaces
arXiv 2607.27222 · 2026-07-31非循环Z-矩阵、树结构矩阵与移位广义超度量矩阵
Inverse acyclic Z-matrices, matrices of tree structure and shifted generalized ultrametric matrices
arXiv 2607.27221 · 2026-07-31非循环矩阵逆中的隐藏衰减
Hidden decay in the inverses of acyclic matrices
arXiv 2607.27220 · 2026-07-31
