2026-08论文中文摘要
可供性即信息：创意媒介作为复杂系统
The Affordance is the Message: Creative Media as Complex Systems
arXiv 2608.12349 · 2026-08-14DCBA：利用诱饵过程检测连通支配集中的协同黑洞攻击
DCBA: Detection of Collaborative Black-Hole Attacks in Connected Dominated Set using Baiting Process
arXiv 2608.12347 · 2026-08-14立场：对齐社区正在无意之中构建审查者的工具包
Position: The Alignment Community is Unintentionally Building a Censor's Toolkit
arXiv 2608.12346 · 2026-08-14评估大型语言模型作为合作科学家的研究完整性的诊断基础
Diagnostic Foundation for Evaluating LLMs' Research Integrity as Co-Scientists
arXiv 2608.12345 · 2026-08-14基于扩散历史预测消费技术拥有情况
Predicting consumer-technology ownership without a diffusion history
arXiv 2608.12344 · 2026-08-14大语言模型是可靠的评审者吗？面向金融文档错误检测的基准测试
Are Large Language Models Reliable Reviewers? A Benchmark for Error Detection in Financial Documents
arXiv 2608.12342 · 2026-08-14大语言模型在文化禁忌安全方面的“知识-行为差距”
The "Knowledge-Behavior Gap" in Cultural Taboo Safety of Large Language Models
arXiv 2608.12341 · 2026-08-14类结构保留优于多样性：不平衡文本分类的文本增强方法综合基准测试
Class-Structure Preservation Beats Diversity: A Comprehensive Benchmark of Text Augmentation Methods for Imbalanced Text Classification
arXiv 2608.12340 · 2026-08-14无理解的模仿：大语言模型中决策偏差的起源
Mimicry without understanding: the origins of decision bias in large language models
arXiv 2608.12339 · 2026-08-14SDAM：面向复杂文本转SQL的结构差异感知记忆演化
SDAM: Structure-Difference-Aware Memory Evolution for Complex Text-to-SQL
arXiv 2608.12338 · 2026-08-14从拒答到丰富：用于长文本幻觉强化学习的评分规则奖励
From Refuse to Richness: Rubric Rewards for Long-Form Hallucination Reinforcement Learning
arXiv 2608.12337 · 2026-08-14StorySpark：面向故事前提生成的模块级进化搜索
StorySpark: Module-wise Evolutionary Search for Story Premise Generation
arXiv 2608.12336 · 2026-08-14HC-RAG：面向异构金融文件的以证据为中心的检索增强生成
HC-RAG: Evidence-Centric Retrieval-Augmented Generation over Heterogeneous Financial Filings
arXiv 2608.12335 · 2026-08-14操控语言轴：从线性可解码到因果控制
Steering the Language Axis: From Linear Decodability to Causal Control
arXiv 2608.12334 · 2026-08-14视觉-语言模型是脆弱的多语言关联器
Vision-Language Models are Fragile Multilingual Associators
arXiv 2608.12333 · 2026-08-14频谱裁剪能否在低秩适配中实现更好的学习同时减少遗忘？
Can Spectral-Clipping Enable Better Learning While Forgetting Less for Low-Rank Adaptation?
arXiv 2608.12332 · 2026-08-14基于自适应注意力匹配的思维感知KV缓存压缩方法用于推理
Thought-Aware KV Cache Compaction for Reasoning via Adaptive Attention Matching
arXiv 2608.12331 · 2026-08-14感知可靠性的性别歧视检测：将DPO与标注者一致性及词元级置信度评分相结合
Reliability-Aware Sexism Detection: Combining DPO with Annotator Agreement and Token-Level Confidence Scoring
arXiv 2608.12330 · 2026-08-14AnchorSIPS：用于证据支持的精神病风险症状测量的合成数据集与评估资源
AnchorSIPS: A Synthetic Dataset and Evaluation Resource for Evidence-Supported Psychosis-Risk Symptom Measurement
arXiv 2608.12329 · 2026-08-14LoRA-Diffusion：基于低秩轨迹分解的参数高效微调方法
LoRA-Diffusion: Parameter-Efficient Fine-Tuning via Low-Rank Trajectory Decomposition
arXiv 2608.12328 · 2026-08-14用于尼泊尔语自动语音识别的多语言预训练模型对比分析
Comparative Analysis of Multilingual Pre-trained Models for Nepali Automatic Speech Recognition
arXiv 2608.12327 · 2026-08-14论法律本体学习中的语义保留度量
On Measuring Semantic Preservation in Legal Ontology Learning
arXiv 2608.12326 · 2026-08-14立场：推理是一个可学习的基于规则的过程
Position: Reasoning is a Learnable Rule-Based Process
arXiv 2608.12325 · 2026-08-14当AI成为你的牧师：大型语言模型的神学分诊与牧灵指导基准测试
When AI Is Your Pastor: A Benchmark for Theological Triage and Pastoral Guidance in Large Language Models
arXiv 2608.12324 · 2026-08-14什么驱动了大语言模型（LLM）的自我反思？武装冲突预测中不确定性路由的受控消融研究
What Drives LLM Self-Reflection? A Controlled Ablation of Uncertainty Routing in Armed Conflict Forecasting
arXiv 2608.12322 · 2026-08-14大型语言模型（LLMs）知道约束但不使用它：语用约束推理中的激活瓶颈
LLMs Know the Constraint But Do Not Use It: Activation Bottlenecks in Pragmatic Constraint Reasoning
arXiv 2608.12321 · 2026-08-14XYZFlow：用于高效生成建模的多维捷径流缩放方法
XYZFlow:Scaling Multi dimensional Shortcut Flows for Efficient Generative Modeling
arXiv 2608.12276 · 2026-08-14信息丰度悖论：长上下文训练会破坏参数化知识
Information Abundance Paradox: Long-Context Training Undermines Parametric Knowledge
arXiv 2608.12218 · 2026-08-14RoutePack：面向MoE强化学习的专家放置与注意力感知数据打包
RoutePack: Expert Placement and Attention-Aware Data Packing for MoE Reinforcement Learning
arXiv 2608.12146 · 2026-08-14超越参数空间：面向鲁棒联邦学习的NTK引导个性化聚合
Beyond Parameter Space: NTK-Guided Personalized Aggregation for Robust Federated Learning
arXiv 2608.12108 · 2026-08-14有色偏紧致哈密顿环的层状障碍
Layer barriers for colour-biased tight Hamilton cycles
arXiv 2608.12013 · 2026-08-14Sci-Surf：通过人类反馈与智能摘要实现科学文献发现
Sci-Surf: Navigating Scientific Literature Discovery through Human Feedback and Intelligent Summarization
arXiv 2608.11973 · 2026-08-14人工智能引发的哲学眩晕
Philosophical vertigo with artificial intelligence
arXiv 2608.11955 · 2026-08-14OEIS Open：语言模型能将多少个猜想转化为定理？
OEIS Open: How many conjectures can language models turn into theorems?
arXiv 2608.11941 · 2026-08-14关于$K^{\boldsymbol{\u221e}}$的注记
Notes on $K^{\infty}$
arXiv 2608.11893 · 2026-08-14UniSwap：用于说话视频的流音频-视觉身份交换
UniSwap: Streaming Audio-Visual Identity Swapping for Talking Videos
arXiv 2608.11752 · 2026-08-14LiveAnimate：稳定的长时长流驱动人体动画实时生成系统
LiveAnimate: Stable Long-Form Streaming Human Animation in Real-Time
arXiv 2608.11745 · 2026-08-14有限维$C^*$代数的Cartan包含的完全刻画
A Complete Characterization of Cartan Inclusions of Finite Dimensional $C^*$-algebras
arXiv 2608.11711 · 2026-08-14面向结构化数据搜索的引导式表格检索
Guided Table Retrieval for Structured Data Search
arXiv 2608.11644 · 2026-08-14云应用向本地部署环境的反向迁移
Reverse Migration of Cloud Applications to On-premises
arXiv 2608.11640 · 2026-08-14MBA：面向现实世界商业创意的多模态基准与智能体
MBA: Multimodal Benchmark and Agents for Real-World Business Ideation
arXiv 2608.11616 · 2026-08-14CookVoice：风格可控的多模态人声生成统一框架
CookVoice: Unified Framework for Style Controllable Multi-Modal Human Voice Generation
arXiv 2608.11590 · 2026-08-14透过梵高的眼睛：基于扩散模型的全局风格迁移
Through Van Gogh's Eyes: Global Style Transfer with Diffusion Model
arXiv 2608.11546 · 2026-08-14部分可观测下循环智能体与脉冲智能体的预测性 allo-static 组织
Predictive Allostatic Organization in Recurrent and Spiking Agents Under Partial Observability
arXiv 2608.11506 · 2026-08-14再访钇超氢化物：YH₆、YH₉与YH₁₀的先进实验与理论研究
Yttrium Superhydrides Revisited: Advanced Experimental and Theoretical Studies of YH$_6$, YH$_9$ and YH$_{10}$
arXiv 2608.11428 · 2026-08-14单循环智能体自摘要下的AI安全护栏存续性
AI Guardrail Survival under Single-Cycle Agentic Self-Summarization
arXiv 2608.11392 · 2026-08-14PAIR：用于RLVR中自适应rollout分配的成对感知包含重加权方法
PAIR: Pairwise-Aware Inclusion Reweighting for Adaptive Rollout Allocation in RLVR
arXiv 2608.11368 · 2026-08-14行动胜于言语：测量使用工具的智能体的跨语言策略保留率
Actions Speak Louder than Words: Measuring Cross-Lingual Policy Retention in Tool-Using Agents
arXiv 2608.11110 · 2026-08-14混合向错-位错系统中的平衡与相互作用 regime
Equilibrium and Interaction Regimes in Mixed Disclination-Dislocation Systems
arXiv 2608.10930 · 2026-08-14光记忆输运成像：面向有限示踪粒子的输运历史框架
Optical-Memory Transport Imaging: A Transport-History Framework for Finite-Memory Tracers
arXiv 2608.10820 · 2026-08-14
