cs.AI（2026-02-19）

📊 共 31 篇论文 | 🔗 1 篇有代码

🎯 兴趣领域导航

支柱九：具身大模型 (Embodied Foundation Models) (22) 支柱二：RL算法与架构 (RL & Architecture) (6 🔗1) 支柱八：物理动画 (Physics-based Animation) (1) 支柱一：机器人控制 (Robot Control) (1) 支柱三：空间感知与语义 (Perception & Semantics) (1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (22 篇)

#题目一句话要点标签🔗⭐
1 What Breaks Embodied AI Security:LLM Vulnerabilities, CPS Flaws,or Something Else? 具身智能安全挑战：超越LLM和CPS，关注系统级匹配问题 embodied AI large language model
2 Evaluating Chain-of-Thought Reasoning through Reusability and Verifiability 提出可复用性和可验证性指标，评估CoT推理质量，揭示现有评估盲点 chain-of-thought
3 Systematic Evaluation of Single-Cell Foundation Model Interpretability Reveals Attention Captures Co-Expression Rather Than Unique Regulatory Signal 系统性评估单细胞Foundation模型可解释性，揭示Attention机制捕获共表达而非独特调控信号 foundation model
4 Enhancing Large Language Models (LLMs) for Telecom using Dynamic Knowledge Graphs and Explainable Retrieval-Augmented Generation 提出KG-RAG框架，利用动态知识图谱增强LLM在电信领域的应用。 large language model
5 A Privacy by Design Framework for Large Language Model-Based Applications for Children 提出面向儿童的LLM应用隐私保护框架，遵循PbD原则与多项法规 large language model
6 Robustness and Reasoning Fidelity of Large Language Models in Long-Context Code Question Answering 研究长文本代码问答中大语言模型的鲁棒性和推理保真度 large language model
7 JEPA-DNA: Grounding Genomic Foundation Models through Joint-Embedding Predictive Architectures JEPA-DNA：通过联合嵌入预测架构实现基因组基础模型的潜在语义对齐 foundation model
8 Decoding the Human Factor: High Fidelity Behavioral Prediction for Strategic Foresight 提出LBM，通过行为嵌入实现高保真个体行为预测，应用于战略预判。 large language model foundation model
9 Agentic Wireless Communication for 6G: Intent-Aware and Continuously Evolving Physical-Layer Intelligence 提出基于LLM的Agentic无线通信框架，实现意图感知和持续演进的物理层智能 large language model multimodal
10 ODESteer: A Unified ODE-Based Steering Framework for LLM Alignment ODESteer：基于ODE的统一框架，用于大语言模型对齐。 large language model
11 Toward a Fully Autonomous, AI-Native Particle Accelerator 提出AI原生粒子加速器设计方案，实现完全自主运行与性能优化 multimodal
12 What Do LLMs Associate with Your Name? A Human-Centered Black-Box Audit of Personal Data 提出LMP2工具，审计大型语言模型中个人数据的关联情况，揭示隐私风险。 large language model
13 Dataless Weight Disentanglement in Task Arithmetic via Kronecker-Factored Approximate Curvature 提出基于K-FAC的无数据权重解耦方法，解决任务算术中的跨任务干扰问题 foundation model
14 ArXiv-to-Model: A Practical Study of Scientific LM Training 针对科学领域，提出一种基于ArXiv LaTeX源码训练领域语言模型的实践方案。 large language model
15 Web Verbs: Typed Abstractions for Reliable Task Composition on the Agentic Web 提出Web Verbs，为Agentic Web构建可靠任务组合的类型化抽象。 large language model
16 Mechanistic Interpretability of Cognitive Complexity in LLMs via Linear Probing using Bloom's Taxonomy 利用Bloom分类学线性探测LLM认知复杂度的机制可解释性 large language model
17 Universal Fine-Grained Symmetry Inference and Enforcement for Rigorous Crystal Structure Prediction 提出通用精细对称性推断与执行框架，用于可靠的晶体结构预测 large language model
18 Instructor-Aligned Knowledge Graphs for Personalized Learning 提出InstructKG框架，构建与教师教学对齐的知识图谱，用于个性化学习。 large language model
19 Toward Trustworthy Evaluation of Sustainability Rating Methodologies: A Human-AI Collaborative Framework for Benchmark Dataset Construction 提出STRIDE和SR-Delta框架，构建可持续性评级基准数据集，提升评级方法的可信度。 large language model
20 How AI Coding Agents Communicate: A Study of Pull Request Description Characteristics and Human Review Responses 研究AI代码生成代理的PR描述特征及人类评审反馈，揭示人机协作软件开发中的交互模式。 large language model
21 Wink: Recovering from Misbehaviors in Coding Agents Wink：用于自动修复自主编码Agent行为异常的轻量级自干预系统 large language model
22 Exploring LLMs for User Story Extraction from Mockups 探索大型语言模型从模型草图中提取用户故事 large language model

🔬 支柱二：RL算法与架构 (RL & Architecture) (6 篇)

#题目一句话要点标签🔗⭐
23 Deep Reinforcement Learning for Optimal Portfolio Allocation: A Comparative Study with Mean-Variance Optimization 提出基于深度强化学习的投资组合优化方法，性能优于传统均值-方差优化。 reinforcement learning deep reinforcement learning DRL
24 Improving LLM-based Recommendation with Self-Hard Negatives from Intermediate Layers 提出ILRec框架，利用中间层自挖掘负样本提升LLM推荐性能 preference learning distillation large language model
25 Retaining Suboptimal Actions to Follow Shifting Optima in Multi-Agent Reinforcement Learning 提出S2Q算法，解决MARL中价值函数漂移导致策略次优问题 reinforcement learning ✅
26 Phase-Aware Mixture of Experts for Agentic Reinforcement Learning 提出Phase-Aware MoE，解决Agent强化学习中的简单任务主导问题 reinforcement learning
27 WebFAQ 2.0: A Multilingual QA Dataset with Mined Hard Negatives for Dense Retrieval 发布WebFAQ 2.0：一个包含挖掘的难负例的多语言QA数据集，用于密集检索。 contrastive learning distillation
28 Continual learning and refinement of causal models through dynamic predicate invention 提出基于动态谓词发明的持续学习框架，用于构建和优化因果模型 PPO world model

🔬 支柱八：物理动画 (Physics-based Animation) (1 篇)

#题目一句话要点标签🔗⭐
29 Deeper detection limits in astronomical imaging using self-supervised spatiotemporal denoising ASTERIS：自监督时空去噪提升天文图像探测极限 spatiotemporal

🔬 支柱一：机器人控制 (Robot Control) (1 篇)

#题目一句话要点标签🔗⭐
30 The Bots of Persuasion: Examining How Conversational Agents' Linguistic Expressions of Personality Affect User Perceptions and Decisions 研究对话Agent人格化语言表达对用户感知和决策的影响 manipulation large language model

🔬 支柱三：空间感知与语义 (Perception & Semantics) (1 篇)

#题目一句话要点标签🔗⭐
31 Visual Model Checking: Graph-Based Inference of Visual Routines for Image Retrieval 提出一种基于图推理视觉程序的图像检索方法，结合形式验证与深度学习。 open-vocabulary open vocabulary

⬅️ 返回 cs.AI 首页 · 🏠 返回主页