2026-09论文中文摘要
DraftAttention2：基于低分辨率引导混合精度注意力的快速视频扩散
DraftAttention2: Fast Video Diffusion with Low-Resolution-Guided Mixed-Precision Attention
arXiv 2609.32628 · 2026-09-29更少量子比特，更好选择：面向量子辅助交通区域划分的耦合感知子QUBO选择
Fewer Qubits, Better Choices: Coupling-Aware Sub-QUBO Selection for Quantum-Assisted Traffic Zone Partitioning
arXiv 2609.32627 · 2026-09-29World SLAM Model：用于SLAM与导航的联合世界建模
World SLAM Model: Joint World Modeling for SLAM and Navigation
arXiv 2609.32626 · 2026-09-29MixDetect：AI编辑的词级定位与量化
MixDetect: Word-Level Localization and Quantification of AI Editing
arXiv 2609.32625 · 2026-09-29五夸克束缚态景观的神经网络变分蒙特卡洛研究
Landscape of pentaquark bound states with neural-network Variational Monte Carlo
arXiv 2609.32624 · 2026-09-29四元数 Hansen--Pedersen 不等式及其在 $S_r$-变换中的应用
A Quaternionic Hansen--Pedersen Inequality and its Application to $S_r$-Transforms
arXiv 2609.32623 · 2026-09-29CoT-Pass@k 真的检验了 CoT 吗？一项多语言数学审计
Does CoT-Pass@k Really Check the CoT? A Multilingual Mathematical Audit
arXiv 2609.32622 · 2026-09-29一种面向6G低轨卫星对地通信系统的动态射线追踪信道模型
A Novel Dynamic Ray-Tracing Channel Model for 6G LEO Satellite-to-Ground Communication Systems
arXiv 2609.32621 · 2026-09-29空间与情境：有限神经资源的竞争决定海马体中印记细胞的分配
Space versus Context: Competition for limited neural resources determines engram cell allocation in the hippocampus
arXiv 2609.32620 · 2026-09-29直觉向量
Intuition vectors
arXiv 2609.32619 · 2026-09-29Python库中废弃API及其替代品的演化
Evolution of Deprecated APIs and Their Replacements in Python Libraries
arXiv 2609.32618 · 2026-09-29对齐悖论：后训练如何放大语言模型中的高置信幻觉
The Alignment Paradox: How Post-Training Amplifies Confident Hallucinations in Language Models
arXiv 2609.32617 · 2026-09-29“你说得对，让我修复它”：当被错误指责时，LLM智能体如何破坏正确的工作
"You're Right, Let Me Fix It": How LLM Agents Damage Correct Work When Falsely Accused
arXiv 2609.32616 · 2026-09-29稳定动态低秩训练
Stabilizing the Dynamic Low-Rank Training
arXiv 2609.32615 · 2026-09-29受奇偶性限制的过度着色分拆 $k$-元组的同余性质
Congruences for Overcolored Partition $k$-tuples Restricted by Parity of the Parts
arXiv 2609.32614 · 2026-09-29同时学习图与嵌入：面向异配节点分类的与分类器无关的重连方法
Learning the Graph and the Embedding Together: Classifier-Independent Rewiring for Heterophilic Node Classification
arXiv 2609.32613 · 2026-09-29加密货币价格相似性：基于皮尔逊相关与动态时间规整的研究
Crypto Price Similarity: An Investigation Using Pearson Correlation and Dynamic Time Warping
arXiv 2609.32611 · 2026-09-29KV-Lingo：通过蒸馏学习 KV 缓存翻译器
KV-Lingo: Learning KV-Cache Translators with Distillation
arXiv 2609.32610 · 2026-09-29从电磁物理到智能优化：室内钉扎天线系统的确定性建模
From Electromagnetic Physics to Intelligent Optimization: A Deterministic Modeling for Indoor Pinching-Antenna Systems
arXiv 2609.32609 · 2026-09-29广义相对论中的热平衡与宇宙学常数的特殊性质
Thermal equilibrium in general relativity and the special nature of the cosmological constant
arXiv 2609.32608 · 2026-09-29VoxMem：大型音频语言模型中的多模态记忆基准
VoxMem: Benchmarking Multimodal Memory in Large Audio Language Models
arXiv 2609.32607 · 2026-09-29John椭球的差分隐私近似
Differentially Private Approximation of the John Ellipsoid
arXiv 2609.32606 · 2026-09-29稠密自-异联想记忆在噪声通信信道中的应用
Dense auto-hetero associative memories applied to noisy communication channels
arXiv 2609.32605 · 2026-09-29当环境成为界面：自动驾驶车辆中人机交互的多感官环境界面
When the Environment Becomes the Interface: Multisensory Environmental Interfaces for Human-AI Interaction in Autonomous Vehicles
arXiv 2609.32604 · 2026-09-29超越NISQ假设：经典可访问随机预言机模型中的一次性存储器
Beyond NISQ Assumptions: One-time Memory in the Classically Accessible Random-Oracle Model
arXiv 2609.32603 · 2026-09-29AnchorRep：通过表示排斥防御大语言模型的跨模型对抗性迁移
AnchorRep: Defending LLMs Against Cross-Model Adversarial Transfer via Representation Repulsion
arXiv 2609.32602 · 2026-09-29VulContextBench：面向编码智能体的安全上下文检索基准
VulContextBench: A Benchmark for Security Context Retrieval in Coding Agents
arXiv 2609.32601 · 2026-09-29CUA-SWE：当计算机使用智能体遇上可视化软件工程
CUA-SWE: When Computer-Use Agents Meet Visual Software Engineering
arXiv 2609.32600 · 2026-09-29奇异簇上0-循环的Riemann-Roch定理
Riemann-Roch for 0-cycles on a singular variety
arXiv 2609.32599 · 2026-09-29中间长波方程的直接散射变换
The Direct Scattering Transform for the Intermediate Long Wave Equation
arXiv 2609.32598 · 2026-09-29并非每个项都增添新结构：符号回归的Sobolev新颖性
Not Every Term Adds New Structure: Sobolev Novelty for Symbolic Regression
arXiv 2609.32597 · 2026-09-29GAUGE：用于前馈4D跟踪的组别视图不一致性修正
GAUGE: Group-Wise View-Inconsistency Rectification for Feed-Forward 4D Tracking
arXiv 2609.32596 · 2026-09-29RECAST：将视觉语言语义重铸为机器人导航的可执行代价地图
RECAST: Recasting Vision-Language Semantics into an Actionable Cost Map for Robot Navigation
arXiv 2609.32595 · 2026-09-29MA-FPPO：多智能体流预训练策略优化
MA-FPPO: Multi-Agent Flow-Pretrained Policy Optimization
arXiv 2609.32594 · 2026-09-29学习时代：预测误差的时间持续性作为学习信号
Age of Learning: Temporal Persistence of Prediction Errors as a Learning Signal
arXiv 2609.32593 · 2026-09-29SPACE：基于反事实驱逐的稀疏预测吸引子用于流式视频记忆
SPACE: Sparse Predictive Attractor via Counterfactual Eviction for Streaming Video Memory
arXiv 2609.32592 · 2026-09-29检索到但未送达：面向长期智能体的多模态记忆投递
Retrieved but Not Delivered: Multimodal Memory Delivery for Long-Term Agents
arXiv 2609.32590 · 2026-09-29度量余型的散度提升与正则分解
Divergence Liftings and Regular Factorizations for Metric Cotype
arXiv 2609.32589 · 2026-09-29足够宽的带状区域唯一极小化平面周期分数阶周长
Sufficiently wide strips uniquely minimize the planar periodic fractional perimeter
arXiv 2609.32588 · 2026-09-29profinite 块与亏群为 $\mathbb Z_p^n$ 的块的有限性性质
Finiteness properties of profinite blocks and blocks with defect group $\mathbb Z_p^n$
arXiv 2609.32587 · 2026-09-29识别 $D_{s1}(2933)$ 为夸克模型中的 $2P_{1}$ 态
Identification of $D_{s1}(2933)$ as the $2P_{1}$ state in the quark model
arXiv 2609.32586 · 2026-09-29表型结构化群体在均匀竞争下适应度渐近无关
Fitness is asymptotically irrelevant under uniform competition in phenotype-structured populations
arXiv 2609.32585 · 2026-09-29EMIR$^2$：演化感知记忆与意图引导的多轮检索
EMIR$^2$: Evolution-Aware Memory with Intent-Guided Multi-Round Retrieval
arXiv 2609.32584 · 2026-09-29临界Lei-Lin-Gevrey空间中阻尼分数阶Navier-Stokes方程的解
On solutions for damped fractional Navier-Stokes equations in a critical Lei-Lin-Gevrey space
arXiv 2609.32583 · 2026-09-29Cool Embeddings：利用AstroCLIP预测IllustrisTNG和TNG-Cluster中的星系团冷却时间
Cool Embeddings: Predicting Galaxy Cluster Cooling Times in IllustrisTNG and TNG-Cluster with AstroCLIP
arXiv 2609.32582 · 2026-09-29HERO-MoE：具有尺度保持融合的历史专家路由
HERO-MoE: Historical Expert Routing with Scale-Preserving Fusion
arXiv 2609.32581 · 2026-09-29DimPO：使用偏好优化进行注意力降维
DimPO: Dimensionality Reduction for Attention using Preference Optimization
arXiv 2609.32579 · 2026-09-29卵泡发育的结构化群体模型
Structured Population Models for Follicular Development
arXiv 2609.32578 · 2026-09-29分组智能体评分与优势再分配用于代码智能体强化学习
Groupwise Agentic Grading and Advantage Redistribution for Code Agent RL
arXiv 2609.32577 · 2026-09-29面向开放式任务的辅助：基于粒子滤波的目标导向共享自主
Assisting for Open-Ended Tasks: Goal-Oriented Shared Autonomy as a Particle Filter
arXiv 2609.32576 · 2026-09-29
