2026-09论文中文摘要
低权重规范逻辑基来自配对划分码
Low-Weight Canonical Logical Bases from Pair-Partition Codes
arXiv 2609.35601 · 2026-09-29等变重数及吸引不动点处的挠
Equivariant multiplicities and torsion at attractive fixed points
arXiv 2609.35600 · 2026-09-29通过函数牛顿更新学习条件期望算子
Learning Conditional Expectation Operators via Functional Newton Updates
arXiv 2609.35598 · 2026-09-29Jouanolou叶状结构在每个次数下的结构稳定性
Structural stability of the Jouanolou foliations in every degree
arXiv 2609.35597 · 2026-09-29树宽与箱约束二次规划的复杂性
Treewidth and the complexity of box-constrained quadratic programs
arXiv 2609.35595 · 2026-09-29噬菌体Q$\eta$：分子与病毒进化前沿的六十年
Bacteriophage Q$β$: Six Decades at the Frontier of Molecular and Viral Evolution
arXiv 2609.35594 · 2026-09-29ReSS：用于3D视觉Transformer的残差恢复稀疏注意力
ReSS: Residual-Restoring Sparse Attention for 3D Vision Transformers
arXiv 2609.35593 · 2026-09-29基于位置的量子密码学中低度策略的指数下界
Exponential lower bounds for low-degree strategies in position-based quantum cryptography
arXiv 2609.35592 · 2026-09-29高阶能量耗散ALE-SAV有限元方法用于两相Navier--Stokes流
High-order energy diminishing ALE-SAV finite element methods for two-phase Navier--Stokes flow
arXiv 2609.35590 · 2026-09-29具有最低可能指标的球面调和映射及其性质
Harmonic maps from spheres with lowest possible index and their properties
arXiv 2609.35589 · 2026-09-29源保留对齐用于科学PDF中稳健的证据定位
Source-preserving alignment for robust evidence localization in scientific PDFS
arXiv 2609.35588 · 2026-09-29面向CUTLASS内核选择的硬件感知特征
Hardware-Aware Features for CUTLASS Kernel Selection
arXiv 2609.35587 · 2026-09-29IMC-CLINIC：用于模拟存内计算裁剪的耦合损失信息牛顿迭代
IMC-CLINIC: Coupled Loss-Informed Newton Iterations for Clipping in Analog In-Memory Computing
arXiv 2609.35586 · 2026-09-29成对评估在视觉扰动下揭示的信息
What Paired Evaluations Reveal under Visual Perturbations
arXiv 2609.35583 · 2026-09-29周期性趋势与基于物理的多阶段高临界温度氢化物超导体搜索
Periodic Trends and a Physics-Based Multistage Search for High-$T_c$ Hydride Superconductors
arXiv 2609.35582 · 2026-09-29QC-Stark：一个揭示大语言模型在量子计算任务上能力分离的多任务基准
QC-Stark: A Multi-Task Benchmark Revealing Capability Dissociations in LLMs Evaluated on Quantum Computing Tasks
arXiv 2609.35581 · 2026-09-29超越最坏情况的量子查询复杂度
Quantum Query Complexity Beyond the Worst Case
arXiv 2609.35580 · 2026-09-29FactorEngram：面向语言模型的基于基级门控的分解N-gram记忆
FactorEngram: Factorized N-gram Memory with Basis-Level Gating for Language Models
arXiv 2609.35578 · 2026-09-29Heisenberg群中SBV函数全变差的BMO型表示公式
A BMO-type representation formula for the total variation of SBV functions in Heisenberg groups
arXiv 2609.35577 · 2026-09-29共享载体AI病毒：跨LLM智能体的记忆跳跃攻击
Share-Borne AI Virus: Memory-Hopping Attacks Across LLM Agents
arXiv 2609.35576 · 2026-09-29最优传输计划的定量稳定性、强制性（Coercivity）与唯一性
Quantitative Stability, Coercivity and Uniqueness of Optimal Transport Plans
arXiv 2609.35574 · 2026-09-29少即是多：用于高效新视角合成的遗传帧选择
Less Is More: Genetic Frame Selection for Efficient Novel View Synthesis
arXiv 2609.35573 · 2026-09-29表示对齐：基于LLM的逆合成规划中的瓶颈
Representation Alignment as a Bottleneck in LLM-Based Retrosynthesis Planning
arXiv 2609.35571 · 2026-09-29EdgeVLN：面向运行时感知的部署就绪量化视觉语言导航模型
EdgeVLN: Runtime-Aware Deployment Ready Quantized Vision Language Navigation Model
arXiv 2609.35570 · 2026-09-29超越能源：当可持续性维度重塑LLM服务决策时
Beyond Energy: When Sustainability Dimensions Reshape LLM Serving Decisions
arXiv 2609.35569 · 2026-09-29从经验到专长：面向数据稀缺NPU内核合成的采纳感知记忆学习
From Experience to Expertise: Adoption-Aware Memory Learning for Data-Scarce NPU Kernel Synthesis
arXiv 2609.35568 · 2026-09-29唯一等变伪期望与交叉积的广义Powers平均
Unique equivariant pseudo expectation and generalized Powers averaging for crossed product
arXiv 2609.35567 · 2026-09-29AI辅助识别磁有序和斯格明子
AI-Assisted Identification of Magnetic Orders and Skyrmions
arXiv 2609.35566 · 2026-09-29Fukaya 范畴与支撑
Fukaya category with supports
arXiv 2609.35565 · 2026-09-29Almieyar：一个基于文化背景的多方言阿拉伯语语音识别基准
Almieyar: A Culturally Grounded Benchmark for Multi-Dialect Arabic Speech Recognition
arXiv 2609.35564 · 2026-09-29中立二次多项式的后临界集的Lebesgue测度
Lebesgue measure of the postcritical set of neutral quadratic polynomials
arXiv 2609.35563 · 2026-09-29重新审视基于视觉Transformer的危险擒抱检测
Revisiting Risky Tackle Detection with Vision Transformers
arXiv 2609.35562 · 2026-09-29WorldPlay2：在控制与视界中扩展实时交互式世界模型
WorldPlay2: Extending Real-Time Interactive World Models in Control and Horizon
arXiv 2609.35560 · 2026-09-29从搜索到研究：自主量化因子挖掘中的搜索扩展探索
From Search to Research: Exploring Search Scaling in Autonomous Quantitative Factor Mining
arXiv 2609.35559 · 2026-09-29六维中逆相对增益阵列的逐项非负性
Entrywise nonnegativity of the inverse relative gain array in dimension six
arXiv 2609.35558 · 2026-09-29编译器可以读取，智能体则不可以：将研究代码的一部分对编码智能体保密
The Compiler May Read It, the Agent May Not: Keeping Part of a Research Code Away from a Coding Agent
arXiv 2609.35557 · 2026-09-29超高真空拉曼平台用于石墨烯纳米带原位表征
Ultra-high vacuum Raman platform for in situ characterization of graphene nanoribbons
arXiv 2609.35556 · 2026-09-29单纯形扩散模型
Simplex Diffusion Models
arXiv 2609.35553 · 2026-09-29BaRe-Mem：用于稳健且自适应智能体咨询的贝叶斯可靠性记忆
BaRe-Mem: Bayesian Reliability Memory for Robust and Adaptive Agent Consultation
arXiv 2609.35551 · 2026-09-29分裂拟阵的 $g$-多项式非负性
Nonnegativity of the $g$-polynomial of split matroids
arXiv 2609.35550 · 2026-09-29关于$E_2$-数的大簇的注记
A note on large clusters of $E_2$-numbers
arXiv 2609.35548 · 2026-09-29亚线性副本数足以通过泡利测量进行保真度估计
Sublinear Copies Suffice for Fidelity Estimation with Pauli Measurements
arXiv 2609.35547 · 2026-09-29重夸克重组在不同碰撞系统中的研究：介子与重子产生
Heavy Quark Recombination Across Collision Systems: Meson and Baryon production
arXiv 2609.35546 · 2026-09-29图世界模型用于受限流行病政策规划
Graph World Models for Constrained Epidemic Policy Planning
arXiv 2609.35545 · 2026-09-29减少谄媚，增强拒绝？从机制可解释性中汲取的AI安全教训
Less Sycophancy, Stronger Refusal? Lessons for AI Safety from Mechanistic Interpretability
arXiv 2609.35544 · 2026-09-29约化群的自守 $C^*$-代数
Automorphic $C^*$-algebras of reductive groups
arXiv 2609.35543 · 2026-09-29有限左单斜括号的分类
Classification of finite left-simple skew braces
arXiv 2609.35542 · 2026-09-29通过双层优化学习鲁棒性机制
Learning the Robustness Mechanism with Bilevel Optimization
arXiv 2609.35541 · 2026-09-29连续上下文管理
Continuous Context Management
arXiv 2609.35540 · 2026-09-29学习以持久对象状态进行推理以实现视频实例分割
Learning to Reason with Persistent Object States for Video Instance Segmentation
arXiv 2609.35539 · 2026-09-29
