2026-10论文中文摘要
Doc2LoRA 提供科学思想的可解码表示
Doc2LoRA Provides Decodable Representations of Scientific Ideas
arXiv 2609.38374 · 2026-10-01NEXUS：在 $z=5$ 处发现一颗明亮的年轻暂现源
NEXUS: Discovery of a Luminous Young Transient at $z=5$
arXiv 2609.38373 · 2026-10-01自进化框架：以智能体自身为优化器在多任务上的应用
Self-Evolving Harness on Multiple Tasks with the Agent as Its Own Optimizer
arXiv 2609.38372 · 2026-10-01TALK-Dem：痴呆相关沟通模式下的具身任务规划基准
TALK-Dem: Benchmarking Embodied Task Planning under Dementia-Associated Communication Patterns
arXiv 2609.38371 · 2026-10-01具有时间和空间依赖性属性的石墨烯等离激元建模的近似方法
An Approximate Method for Modeling Plasmons on Graphene with Time- and Space-dependent Properties
arXiv 2609.38370 · 2026-10-01AI代理能否重新发现Blaschke曲线不变量？
Can an AI Agent Rediscover a Blaschke-Curve Invariant?
arXiv 2609.38369 · 2026-10-01组合，而非对话：视觉语言模型丢失的是场景，不是线索
Composition, Not Conversation: VLMs Lose the Scene, Not the Thread
arXiv 2609.38368 · 2026-10-01可证明的近似最优性：统一（半）随机CSP搜索与反驳的简单框架
Certifiable Near-Optimality: A Simple Framework for Unifying Search and Refutation for (Semi)random CSPs
arXiv 2609.38367 · 2026-10-01关于有限群Reidemeister谱的一个推广
On a generalization of the Reidemeister spectrum of a finite group
arXiv 2609.38366 · 2026-10-01离散平均生成器加速扩散语言模型
Acceleration of Diffusion Language Model through Discrete Average Generator
arXiv 2609.38364 · 2026-10-01模拟器精化扩散用于射频逆向设计
Simulator-Refined Diffusion for Radio-Frequency Inverse Design
arXiv 2609.38363 · 2026-10-01归纳视觉逻辑用于视觉语言模型的小样本分布外适应
Inductive Visual Logic for Few-Shot Out-Of-Distribution Adaptation in VLMs
arXiv 2609.38362 · 2026-10-01数据驱动亚组发现后标准试验估计量的推断
Inference for Standard Trial Estimands after Data-Driven Subgroup Discovery
arXiv 2609.38361 · 2026-10-01论在线策略蒸馏中的离线策略教师
On the Off-Policy Teacher in On-Policy Distillation
arXiv 2609.38360 · 2026-10-01超越模式坍缩：通过生成流网络生成多样化的合成专家对话
Beyond Mode Collapse: Generating Diverse Synthetic Expert Conversations via Generative Flow Networks
arXiv 2609.38359 · 2026-10-01生物探索动力学的轨迹统计
The Trajectory Statistics of Biological Exploratory Dynamics
arXiv 2609.38358 · 2026-10-01评估语言模型在长对抗性对话中的安全性
Evaluating Language Model Safety Across Long Adversarial Conversations
arXiv 2609.38357 · 2026-10-01循环神经网络中通过可回收单元门控实现动力系统的持续学习
Continual Learning of Dynamical Systems in Recurrent Neural Networks through Recyclable Unit Gating
arXiv 2609.38356 · 2026-10-01时间积分泄漏作为多模超导量子比特复位的动力学基准
Time-Integrated Leakage as a Dynamical Benchmark for Multi-Mode Superconducting-Qubit Reset
arXiv 2609.38354 · 2026-10-01可编程耗散光子电路中动量空间拓扑的干涉测量读取
Interferometric Readout of Momentum-Space Topology in a Programmable Dissipative Photonic Circuit
arXiv 2609.38352 · 2026-10-01尼泊尔Syabrubesi洪水传播的三维模拟
Three-Dimensional Simulation of Flood Propagation at Syabrubesi, Nepal
arXiv 2609.38351 · 2026-10-01AGN SED拟合的新希望：基于C2PO-Torus模型的12MGS AGN X射线光谱分析
A New Hope for AGN SED Fitting: X-Ray Spectral Analysis of 12MGS AGN with the C2PO-Torus Model
arXiv 2609.38350 · 2026-10-01MILO：通过编排式多智能体进化实现自动化工具链发现
MILO: Automated Harness Discovery via Orchestrated Multi-Agent Evolution
arXiv 2609.38349 · 2026-10-01函数空间Transformer与自适应锚点
Function-Space Transformer with Adaptive Anchors
arXiv 2609.38348 · 2026-10-01TrackFish3D：基于多视角视频的鱼群自监督3D跟踪
TrackFish3D: Self-Supervised 3D Tracking of Schooling Fish from Multi-view Videos
arXiv 2609.38347 · 2026-10-01考察引导式AI导师如何解决学生困境的差异
Examining Variation in How Guided AI Tutors Resolve Student Impasses
arXiv 2609.38346 · 2026-10-01VLBI天体测量程序与CHIME/FRB Outriggers的性能
VLBI Astrometric Procedure and Performance with CHIME/FRB Outriggers
arXiv 2609.38344 · 2026-10-01学习可动画高斯头部虚拟形象的面部语义修复
Learning Semantic Inpainting for Animatable Gaussian Head Avatars
arXiv 2609.38343 · 2026-10-01激活条件自蒸馏
Activation-Conditioned Self-Distillation
arXiv 2609.38342 · 2026-10-01HESS J1507-622：一个可能的年轻银河系千新星遗迹
HESS J1507-622: A Plausible Young Galactic Kilonova Remnant
arXiv 2609.38341 · 2026-10-01CARAT：材料大语言模型是在推理还是背诵？
CARAT: Do Materials LLMs Reason or Recite?
arXiv 2609.38340 · 2026-10-01Aegis：面向隐私保护医学联邦学习的生成式梯度掩蔽
Aegis: Generative Gradient Masking for Privacy-Preserving Medical Federated Learning
arXiv 2609.38339 · 2026-10-01在 $L^2$ 可积基中的中子-原子核散射
Neutron-nucleus scattering in an $L^2$-integrable basis
arXiv 2609.38338 · 2026-10-01E2E-SWE：基准测试大语言模型从零构建可用代码库
E2E-SWE: Benchmarking LLMs on Building Working Codebases from Scratch
arXiv 2609.38335 · 2026-10-01自旋链中受约束耗散的集体动力学：经典脆弱但量子鲁棒
Collective Dynamics in Spin Chains with Constrained Dissipation: Classically Fragile but Quantum Robust
arXiv 2609.38333 · 2026-10-01Hermes：学习上下文推理解锁测试时扩展
Hermes: Learning Contextual Reasoning Unlocks Test-Time Scaling
arXiv 2609.38332 · 2026-10-01迈向用物理信息神经网络求解广义相对论
Towards solving General Relativity with Physics-Informed Neural Networks
arXiv 2609.38331 · 2026-10-01通过SDSS测光红移的宇宙邻域关联法估算VHE及候选VHE BL Lac天体的距离
Cosmic-neighbor-associated distances to VHE and candidate VHE BL Lacs via SDSS photometric redshifts
arXiv 2609.38330 · 2026-10-01ExploreNet: 学习在扩散GRPO中何处探索
ExploreNet: Learning Where to Explore in Diffusion GRPO
arXiv 2609.38329 · 2026-10-01盲量子数据压缩的指数强逆定理
Exponential strong converse for blind quantum data compression
arXiv 2609.38328 · 2026-10-01学习Gaia：基于Gaia数据发布3源目录训练的生成模型
Learning Gaia: A Generative Model Trained on the Gaia Data Release 3 Source Catalog
arXiv 2609.38326 · 2026-10-01和弦共鸣！模态动态字体设计
Strike a Chord! Modal Kinetic Typography
arXiv 2609.38325 · 2026-10-01当异议被保留时，多智能体讨论的收益减少
Multi-agent discussion gains less when dissent is withheld
arXiv 2609.38324 · 2026-10-01局部代数的张量分解
Tensor factorization of local algebras
arXiv 2609.38323 · 2026-10-01揭示变状态活动星系核中的吸积盘结构：NGC1566
Revealing the accretion disc structure in Changing State AGN: NGC1566
arXiv 2609.38322 · 2026-10-01X射线极宽视场光谱巡天仪（XESS，或“Chess”）
An X-ray Extremely wide-field Spectroscopic Surveyor (XESS, or "chess")
arXiv 2609.38321 · 2026-10-01锂丰富巨行星中的密近双星巡天
Surveying for Close Binaries among Lithium-rich Giants
arXiv 2609.38320 · 2026-10-01当对称性抑制魔法时
When Symmetry Suppresses Magic
arXiv 2609.38319 · 2026-10-01全息共形界面在一般维度中的能量传输
Energy Transmission Across Holographic Conformal Interfaces in General Dimensions
arXiv 2609.38318 · 2026-10-01对称晶体中的隐藏角动量环流
Hidden Angular Momentum Loop Currents in Symmetric Crystals
arXiv 2609.38317 · 2026-10-01
