2026-10论文中文摘要
多任务主动学习与高效资源分配
Multi-Task Active Learning with Efficient Resource Allocation
arXiv 2610.07045 · 2026-10-07Szykuła 图3族的“与另一压缩”阈值
The compress-with-another threshold of Szykuła's Figure 3 family
arXiv 2610.07044 · 2026-10-07TRIAGE：原生NVFP4强化学习的方向感知失配稳定化
TRIAGE: Direction-Aware Mismatch Stabilization of Native NVFP4 Reinforcement Learning
arXiv 2610.07043 · 2026-10-07XuYi 干涉仪观测模拟器
XuYi Interferometer Observation Simulator
arXiv 2610.07042 · 2026-10-07金砖国家南方天文台（BRICS-SO）项目：邀请非洲国家投资智利
BRICS Southern Observatory (BRICS-SO) Project: Invitation to African countries to invest in Chile
arXiv 2610.07041 · 2026-10-07通过重参数化进行优化：具有奇异几何的时间扭曲镜像流
Optimization by Reparametrization: Time-Warped Mirror Flows with Singular Geometry
arXiv 2610.07040 · 2026-10-07Tellimation：通过即时动画让儿童叙事中的叙事空白可见
Tellimation: Making Narrative Gaps Visible in Children's Storytelling with Just-in-Time Animation
arXiv 2610.07039 · 2026-10-07问题在于前提：自监控测试时自适应中的可交换性失效
The Premise Is the Problem: Exchangeability Failure in Self-Monitored Test-Time Adaptation
arXiv 2610.07038 · 2026-10-07物理有效生物分子扩散模型的推理时投影
Inference-Time Projection for Physically Valid Biomolecular Diffusion Models
arXiv 2610.07037 · 2026-10-07JIVEAdapter：一种基于联合与个体变异解释（JIVE）的多任务加性低秩适配器
JIVEAdapter: A Multi-Task Additive Low-Rank Adapter via Joint and Individual Variation Explained (JIVE)
arXiv 2610.07036 · 2026-10-07IO500基准测试的仓库级性能特征分析
Repository-Scale Performance Characterization of the IO500 Benchmark
arXiv 2610.07035 · 2026-10-07塑造风场：用于运动学可容许城市风预测的嵌套势函数
Shaping the Wind: Nested Potentials for Kinematically Admissible Urban Wind Prediction
arXiv 2610.07033 · 2026-10-07探究生物医学领域神经机器翻译的模型压缩
Investigating Model Compression for Neural Machine Translation in the Biomedical Domain
arXiv 2610.07032 · 2026-10-07Artemis：基于几何约束的多智能体驾驶世界模型，具有共享3D状态与渐进式记忆更新
Artemis: Geometry-Grounded Multi-Agent Driving World Models with Shared 3D State and Progressive Memory Update
arXiv 2610.07031 · 2026-10-07基于LLM的多智能体协作实现约束多目标容器放置
LLM-Based Multi-Agent Collaboration for Constrained Multi-Objective Container Placement
arXiv 2610.07030 · 2026-10-07基于ReRAM的存内计算CNN加速器的经验性故障脆弱性探索
An Empirical Fault Vulnerability Exploration of ReRAM-based Process-in-Memory CNN Accelerators
arXiv 2610.07029 · 2026-10-07来自预训练扩散表示的可识别世界模型
Identifiable World Models from Pretrained Diffusion Representations
arXiv 2610.07028 · 2026-10-07培养未来工程师关于大型语言模型的能力：一个可扩展的工作坊
Educating future engineers about LLMs: A scalable workshop
arXiv 2610.07027 · 2026-10-07离线AI模块：语音优先的离线架构、硬件参考栈、量化与基准测试
Offline AI Modules: Voice-First Offline Architecture, Hardware Reference Stack, Quantization and Benchmarking
arXiv 2610.07026 · 2026-10-07WiSPER：基于WiFi CSI的多人体3D姿态估计的姿势监督预测与残差流细化
WiSPER: Pose-Supervised Predictive and Residual Flow Refinement For Multi-Person 3D Pose Estimation With WiFi CSI
arXiv 2610.07025 · 2026-10-07基于CMUT集成治疗与成像阵列的经颅聚焦超声中无MRI的CT到超声颅骨配准
MRI-Free CT-to-Ultrasound Skull Registration in Transcranial Focused Ultrasound using CMUT based Integrated Therapeutic and Imaging Arrays
arXiv 2610.07024 · 2026-10-07超越拒绝模式：安全角色内化实现鲁棒且可泛化的LLM安全对齐
Beyond Refusal Patterns: Safe-Role Internalization for Robust and Generalizable LLM Safety Alignment
arXiv 2610.07023 · 2026-10-07非零剩余类中的 $kK_3$ 边谱极值问题
The edge spectral extremal problem for $kK_3$ in nonzero residue classes
arXiv 2610.07022 · 2026-10-07基于锥形相位调制的免校准宽带重建光谱仪
Calibration-Free Broadband Reconstructive Spectrometer based on Conical Phase Modulation
arXiv 2610.07021 · 2026-10-07围绕被完美流体暗物质包围的旋转黑洞的周期轨道及其引力波特征
Periodic Orbits and Their Gravitational-Wave Signatures around a Rotating Black Hole Surrounded by Perfect Fluid Dark Matter
arXiv 2610.07020 · 2026-10-07来自40亿参数开放模型的随机试验校准答案：一项注册测试与许可证清洁发布
Calibrated Answers About Randomized Trials From a 4-Billion-Parameter Open Model: A Registered Test and a License-Clean Release
arXiv 2610.07019 · 2026-10-07何时重新思考：为视觉-语言模型学习多视角自验证
When to Rethink: Learning Multi-Perspective Self-Verification for Vision-Language Models
arXiv 2610.07018 · 2026-10-07A$_{\infty}$-范畴范畴中余等化子不存在的一个例子
An example of the non-existence of coequalizers in the category of the A$_{\infty}$-categories
arXiv 2610.07017 · 2026-10-07锚定与适配：少样本工业异常检测的非对称提示适配
Anchor and Adapt: Asymmetric Prompt Adaptation for Few-Shot Industrial Anomaly Detection
arXiv 2610.07016 · 2026-10-07LEAP：使特权几何监督对视觉运动学习有效
LEAP: Making Privileged Geometry Supervision Effective for Visuomotor Learning
arXiv 2610.07015 · 2026-10-07DTFormer：用于RGB-D分割的文本引导语义对齐
DTFormer: Text-Guided Semantic Alignment for RGB-D Segmentation
arXiv 2610.07014 · 2026-10-07热场电子发射计算的数值高效模型
A numerical and efficient model for thermo-field electron emission calculations
arXiv 2610.07013 · 2026-10-07重复双极性方波电压下利兹线绝缘的频率依赖性击穿
Frequency-Dependent Breakdown of Litz-Wire Insulation Under Repetitive Bipolar Square-Wave Voltages
arXiv 2610.07012 · 2026-10-07一种使用极值分布的FAS信道拟合策略以实现精确的中断性能评估
A FAS Channel Fitting Strategy Using Extreme Value Distributions for Accurate Outage Performance Evaluation
arXiv 2610.07011 · 2026-10-07精确酉合成：在尺寸与深度-辅助量子比特权衡上同时最优
Exact Unitary Synthesis: Simultaneously Optimal in Size and Depth-Ancilla Tradeoff
arXiv 2610.07010 · 2026-10-07哪个图像属性携带了越狱？图像到文本越狱的受控剖析
Which Image Property Carries the Jailbreak? A Controlled Dissection of Image-to-Text Jailbreaks
arXiv 2610.07009 · 2026-10-07学习策展生成内容以实现泛化的小样本类增量学习
Learning to Curate What You Generate for Generalizable Few-Shot Class-Incremental Learning
arXiv 2610.07008 · 2026-10-07识别塑性无机半导体需要更严格的标准
Identifying Plastic Inorganic Semiconductors Requires More Rigorous Criteria
arXiv 2610.07007 · 2026-10-07音频越狱存在于何处？对Qwen2-Audio上AdvWave-P的受控频率-解码器深度审计
Where Does the Audio Jailbreak Live? A Controlled Frequency-Depth Audit of AdvWave-P on Qwen2-Audio
arXiv 2610.07005 · 2026-10-07基于细胞工作流复形的拓扑一致任务规划框架用于LLM智能体
Topology-Consistent Task Planning over Cellular Workflow Complexes for LLM-based Agents
arXiv 2610.07004 · 2026-10-07AI智能体的可靠性：评分者效应、漂移与评估项目的回归
Reliability of AI Agents: Rater Effects, Drift, and the Return to an Evaluation Program
arXiv 2610.07003 · 2026-10-07我们应跳过扩散吗？
Should We Skip Diffusion?
arXiv 2610.07002 · 2026-10-07弱型非局部泛函的全范围 $\Gamma$-收敛
Full-Range $Γ$-Convergence of Weak-Type Nonlocal Functionals
arXiv 2610.07001 · 2026-10-07信任门控能力控制：打破多智能体LLM系统中的信任-脆弱性悖论
Trust-Gated Capability Control: Breaking the Trust-Vulnerability Paradox in Multi-Agent LLM Systems
arXiv 2610.07000 · 2026-10-07ProactiveVLA：通过主动环境探索增强具身记忆
ProactiveVLA: Augmenting Embodied Memory through Proactive Environment Exploration
arXiv 2610.06999 · 2026-10-07具有文件和需求隐私性的阶最优编码缓存
Order-Optimal Coded Caching With File and Demand Privacy
arXiv 2610.06998 · 2026-10-07时间不一致停止问题的惩罚方法的局部 $W^{2,1}_{p}$ 收敛性
Local $W^{2,1}_{p}$ convergence of penalization for time-inconsistent stopping problems
arXiv 2610.06997 · 2026-10-07块扩散语言模型中的掩码引导KV缓存驱逐
Mask-Guided KV Cache Eviction in Block Diffusion Language Models
arXiv 2610.06996 · 2026-10-07联合上界覆盖与路径选择效用：基于两个城市代理任务的实证评估
Joint upper-bound coverage and route-choice utility: an empirical evaluation on two urban proxy tasks
arXiv 2610.06995 · 2026-10-07TARE：在阅读后门防御成本之前，先称量一个从未被投毒的孪生模型
TARE: Weigh a Never-Poisoned Twin Before Reading Backdoor-Defense Costs
arXiv 2610.06994 · 2026-10-07
