2026-10论文中文摘要
评估用于LLM路由的升级信号：目标、控制以及五种自欺方式
Evaluating Escalation Signals for LLM Routing: Targets, Controls, and Five Ways to Fool Yourself
arXiv 2610.07354 · 2026-10-07有限磁场下QCD相图与临界端点：基于神经网络的全息方法
The QCD phase diagram and critical end point at finite magnetic field from holography with neural networks
arXiv 2610.07353 · 2026-10-07Krylov空间中的经典混沌特征
Classical chaotic signatures in Krylov space
arXiv 2610.07352 · 2026-10-07基于小指数哈希的简单极强有损函数
Simple Extremely Lossy Functions from Small-Exponent Hashing
arXiv 2610.07351 · 2026-10-07轨迹检索投机解码：模型自身历史何时有帮助？
Trajectory-Retrieval Speculative Decoding: When Does a Model's Own History Help?
arXiv 2610.07350 · 2026-10-07RELACE：面向长时程语言智能体的基于回溯似然的动作信用估计
RELACE: retrospective likelihood-based action credit estimation for long-horizon language agents
arXiv 2610.07349 · 2026-10-07阶梯式专家混合：具有可配置推理复杂度的分段级路由
Stepped MoE: Segment-Level Routing with Configurable Inference Complexity
arXiv 2610.07348 · 2026-10-07BCZ 映射的混合性
Mixing of the BCZ Map
arXiv 2610.07347 · 2026-10-07ScanSTL：信号时序逻辑的并行鲁棒性评估
ScanSTL: Parallel Robustness Evaluation for Signal Temporal Logic
arXiv 2610.07346 · 2026-10-07评估云环境中可解释IAM策略风险评分的行为上下文
Evaluating Behavioral Context for Interpretable IAM Policy Risk Scoring in Cloud Environments
arXiv 2610.07345 · 2026-10-07极端分位数回归中加权最大似然估计量的存在性与一致性
Existence and consistency of weighted maximum likelihood estimator for extreme quantile regression
arXiv 2610.07344 · 2026-10-07理性引导的策略优化：利用自适应理性脚手架学习推理
Rationale-Guided Policy Optimization: Learning to Reason with Adaptive Rationale Scaffolding
arXiv 2610.07342 · 2026-10-07具有有向相互作用的线性动力学反应-扩散系统从单快照二阶统计量的可辨识性
Identifiability of a linear dynamical reaction-diffusion system with directed interactions from the second-order statistics of a single snapshot
arXiv 2610.07341 · 2026-10-07CausalBind：蛋白质-分子虚拟筛选的因果建模与学习
CausalBind: Causal Modeling and Learning for Protein-Molecule Virtual Screening
arXiv 2610.07340 · 2026-10-07战术战伤救护的教义基础视觉问答数据集
A doctrine-grounded visual question answering dataset for Tactical Combat Casualty Care
arXiv 2610.07339 · 2026-10-07Logbook：极长时音频事件理解
Logbook: Extremely Long-form Audio Event Understanding
arXiv 2610.07338 · 2026-10-07基于层级接地语义手术的文本到图像扩散模型中的组合概念擦除
Compositional Concept Erasure in Text-to-Image Diffusion Models via Hierarchically Grounded Semantic Surgery
arXiv 2610.07337 · 2026-10-07类量子空间决策动力学：沉浸式导航中线索顺序效应的可证伪模型及其对人-量子计算机交互的启示
Quantum-Like Spatial Decision Dynamics: A Falsifiable Model of Cue-Order Effects in Immersive Navigation with Implications for Human-Quantum Computer Interaction
arXiv 2610.07336 · 2026-10-07选择性批判：面向长时程决策中成本感知的LLM智能体
Selective Critique for Cost-Aware LLM Agents in Long-Horizon Decision Making
arXiv 2610.07335 · 2026-10-07权重预言机：用语言模型读取神经网络权重
Weight Oracles: Reading Neural Network Weights with Language Models
arXiv 2610.07334 · 2026-10-07面向分布式MoE训练的内存高效专家路由
Memory-Efficient Expert Routing for Distributed MoE Training
arXiv 2610.07333 · 2026-10-07结构化MoE专家选择以支持智能体强化学习
Structuring MoE Expert Selection for Agentic Reinforcement Learning
arXiv 2610.07332 · 2026-10-07旁观者暗物质与希格斯门户
Spectator Dark Matter and the Higgs Portal
arXiv 2610.07331 · 2026-10-07关于 $n$ 个同构群之间的弱交换性，$n \geq 2$
On weak commutativity between $n$ isomorphic groups, $n \geq 2$
arXiv 2610.07330 · 2026-10-07大型科研团队更具颠覆性，而非更不具颠覆性
Large scientific teams are more, not less, disruptive
arXiv 2610.07329 · 2026-10-07量子测量类学习中纠缠学习规则的优势
Advantage of Entangled Learning Rules in Quantum Measurement Class Learning
arXiv 2610.07328 · 2026-10-07SharedKV-BT：行为树智能体的节点局部类型化决策
SharedKV-BT: Node-Local Typed Decisions for Behavior-Tree Agents
arXiv 2610.07327 · 2026-10-07无需人工标注的X射线安检扫描中任意物体定位
Localize Any Object in X-Ray Security Scans without Human Annotation
arXiv 2610.07326 · 2026-10-07快速射电暴宿主星系金属丰度的空间分辨视角：积分场光谱研究
A spatially resolved view of Fast Radio Burst host metallicities with Integral-field Spectroscopy
arXiv 2610.07325 · 2026-10-07时间序列基础模型的尺度不变训练
Scale-Invariant Training for Time Series Foundation Models
arXiv 2610.07324 · 2026-10-07ATLAS-AL：基于主动学习的潜在对抗搜索自适应信赖域
ATLAS-AL: Adaptive Trust-Region for Latent Adversarial Searches via Active Learning
arXiv 2610.07323 · 2026-10-07正温等离子体柱中正电子加速能量回收的极限
Limits on energy recovery for positron acceleration in a warm plasma column
arXiv 2610.07322 · 2026-10-07BrainScaleS 晶圆级神经形态系统的突触损失估计
Synapse Loss Estimation for the BrainScaleS Wafer-scale Neuromorphic System
arXiv 2610.07321 · 2026-10-07面向机器人群体最小长度通信网络的去中心化构建
Towards Decentralized Formation of Minimum-Length Communication Networks Using Robot Swarms
arXiv 2610.07320 · 2026-10-07类乐高式平面柔顺模块刚度配置方法用于任务特定柔性接口
Lego-Like Stiffness Configuration of Planar Compliant Modules for Task-Specific Flexible Interfaces
arXiv 2610.07319 · 2026-10-07关于两个拟阵的列表色数的一个注记
A note on the list chromatic number of two matroids
arXiv 2610.07318 · 2026-10-07Rudolph-Johnson 无动力学：圆周上的自同态刚性
Rudolph-Johnson without dynamics: endomorphic rigidity on the circle
arXiv 2610.07317 · 2026-10-07过渡金属二硫族化物单层中共振拉曼谱的理论：谷间声子散射引起的干涉效应
Theory of resonance Raman profiles in transition metal dichalcogenide monolayers: Interference effects due to inter-valley phonon scattering
arXiv 2610.07316 · 2026-10-07除尘装置中粉尘颗粒的运动方程（译自苏联期刊《供暖与通风》1934年第4期第27-29页）
Equations of motion for dust particles in dust-collecting devices (Translated article from USSR Journal Otoplenie i Ventilyatsiya, No. 4 (1934): 27-29)
arXiv 2610.07315 · 2026-10-07基于广义等角测量的维度相关导引见证的解析构造
Analytic construction of dimension-dependent steering witnesses from generalized equiangular measurements
arXiv 2610.07314 · 2026-10-07基于规则的神经符号AI语言
Rule-Based Languages for Neurosymbolic AI
arXiv 2610.07313 · 2026-10-07通过LHC上的伴随规范玻色子反冲和极化来探测轴子类粒子
Probing axion-like particles through associated gauge-boson recoil and polarization at the LHC
arXiv 2610.07312 · 2026-10-07理解并利用智能体记忆缓解推理时的过度依赖
Understanding and Mitigating Inference-Time Overreliance Using Agentic Memory
arXiv 2610.07311 · 2026-10-07从沙箱到执行：面向关键基础设施的置信度限定威胁情报
From Sandbox to Enforcement: Confidence-Qualified Threat Intelligence for Critical Infrastructure
arXiv 2610.07310 · 2026-10-07错误语境中的正确记忆：验证长期智能体记忆中的检索可接纳性
The Right Memory in the Wrong Context: Verifying Retrieval Admissibility in Long-Term Agent Memory
arXiv 2610.07309 · 2026-10-07量子动力学：利用量子类资源模拟自旋系统
Quantum dynamics of spin systems using quantum-like resources
arXiv 2610.07308 · 2026-10-07非合作可微博弈中的基于度量的均衡选择
Metric-Based Equilibrium Selection in Noncooperative Differentiable Games
arXiv 2610.07307 · 2026-10-07Kurate：可扩展的科学质量分析
Kurate: Scalable Scientific Quality Analysis
arXiv 2610.07306 · 2026-10-07小密度差，大理查德森数：浮力决定石膏在离心微流控汇流单元中的沉积位置
Small density differences, large Richardson numbers: buoyancy decides where gypsum deposits in a centifluidic confluence cell
arXiv 2610.07305 · 2026-10-07无仿星器对称性的解析非轴对称三维磁流体力学平衡
Analytic non-axisymmetric 3D MHD equilibria without stellarator symmetry
arXiv 2610.07304 · 2026-10-07
