cs.AI（2025-10-07）

📊 共 35 篇论文 | 🔗 4 篇有代码

🎯 兴趣领域导航

支柱九：具身大模型 (Embodied Foundation Models) (26 🔗3) 支柱二：RL算法与架构 (RL & Architecture) (6) 支柱一：机器人控制 (Robot Control) (1 🔗1) 支柱三：空间感知与语义 (Perception & Semantics) (1) 支柱四：生成式动作 (Generative Motion) (1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (26 篇)

#题目一句话要点标签🔗⭐
1 MetaVLA: Unified Meta Co-training For Efficient Embodied Adaption MetaVLA：用于高效具身适应的统一元协同训练框架 vision-language-action VLA OpenVLA
2 ARM: Discovering Agentic Reasoning Modules for Generalizable Multi-Agent Systems 提出ARM：发现通用多智能体系统的Agentic推理模块 large language model foundation model chain-of-thought
3 BuilderBench -- A benchmark for generalist agents BuilderBench：面向通用智能体，用于开放式探索的基准测试平台 generalist agent
4 Domain-Shift-Aware Conformal Prediction for Large Language Models 提出领域偏移感知共形预测(DS-CP)，提升大语言模型在领域偏移下的不确定性量化 large language model
5 PuzzlePlex: Benchmarking Foundation Models on Reasoning and Planning with Puzzles PuzzlePlex：用于评估具身智能体推理与规划能力的多样化谜题基准 foundation model
6 Leveraging Large Language Models for Cybersecurity Risk Assessment -- A Case from Forestry Cyber-Physical Systems 利用本地化大语言模型辅助林业网络物理系统网络安全风险评估 large language model
7 StarEmbed: Benchmarking Time Series Foundation Models on Astronomical Observations of Variable Stars StarEmbed：天文学变星观测时间序列基础模型基准测试 foundation model
8 Digital Transformation Chatbot (DTchatbot): Integrating Large Language Model-based Chatbot in Acquiring Digital Transformation Needs 提出基于大语言模型的数字化转型需求获取聊天机器人DTchatbot large language model
9 Early Multimodal Prediction of Cross-Lingual Meme Virality on Reddit: A Time-Window Analysis 提出一种基于时间窗口分析的跨语言Meme早期流行度多模态预测方法 multimodal
10 Uncovering Representation Bias for Investment Decisions in Open-Source Large Language Models 揭示开源大语言模型在投资决策中的表征偏差，关注Qwen模型 large language model
11 Membership Inference Attacks on Tokenizers of Large Language Models 提出基于Tokenizer的成员推断攻击，揭示大语言模型隐私风险 large language model
12 Data Provenance Auditing of Fine-Tuned Large Language Models with a Text-Preserving Technique 提出一种文本保持的水印框架，用于审计微调大语言模型的数据来源 large language model
13 Large Language Model-Based Uncertainty-Adjusted Label Extraction for Artificial Intelligence Model Development in Upper Extremity Radiography GPT-4o提取放射报告标签，用于上肢X光片多标签图像分类模型训练 large language model
14 Moloch's Bargain: Emergent Misalignment When LLMs Compete for Audiences 揭示LLM竞争中涌现的“莫洛克交易”：追求成功导致AI对齐性下降 large language model
15 Domain-Grounded Evaluation of LLMs in International Student Knowledge 针对留学知识领域，提出领域相关的LLM评估方法，解决幻觉问题。 large language model
16 Relative Positioning Based Code Chunking Method For Rich Context Retrieval In Repository Level Code Completion Task With Code Language Model 提出基于相对位置的代码块划分方法，提升代码语言模型在仓库级代码补全任务中的性能 large language model
17 Orders in Chaos: Enhancing Large-Scale MoE LLM Serving with Data Movement Forecasting 通过数据移动预测增强大规模MoE LLM Serving性能 large language model ✅
18 Impact of LLMs on Team Collaboration in Software Development 研究LLM对软件开发团队协作的影响，提升效率与沟通，应对挑战与安全问题。 large language model
19 Automated Program Repair of Uncompilable Student Code 利用大型语言模型自动修复学生未编译代码，提升学生建模效果 large language model
20 MixReasoning: Switching Modes to Think MixReasoning：提出一种自适应调整推理深度的混合推理框架 chain-of-thought
21 Training-Free Time Series Classification via In-Context Reasoning with LLM Agents 提出FETA：基于LLM Agent上下文推理的免训练时间序列分类框架 large language model ✅
22 Optimizing for Persuasion Improves LLM Generalization: Evidence from Quality-Diversity Evolution of Debate Strategies DebateQD：基于说服力优化的LLM提升泛化能力，解决过拟合问题 large language model
23 VeriEquivBench: An Equivalence Score for Ground-Truth-Free Evaluation of Formally Verifiable Code 提出VeriEquivBench基准，用于无ground-truth评估形式化可验证代码的等价性。 large language model
24 ConstraintLLM: A Neuro-Symbolic Framework for Industrial-Level Constraint Programming 提出ConstraintLLM以解决工业级约束编程问题 large language model ✅
25 Artificially intelligent agents in the social and behavioral sciences: A history and outlook 回顾社会与行为科学中智能代理的发展历程与未来展望 large language model
26 From Agentification to Self-Evolving Agentic AI for Wireless Networks: Concepts, Approaches, and Future Research Directions 提出自进化Agentic AI框架，解决无线网络中人工干预的优化难题 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🔗⭐
27 The Safety Challenge of World Models for Embodied AI Agents: A Review 综述世界模型在具身智能安全挑战，分析自动驾驶与机器人场景下的模型缺陷 world model embodied AI
28 Joint Communication Scheduling and Velocity Control for Multi-UAV-Assisted Post-Disaster Monitoring: An Attention-Based In-Context Learning Approach 提出基于注意力机制的上下文学习方法AIC-VDS，用于多无人机辅助的灾后监测通信调度与速度控制联合优化。 reinforcement learning deep reinforcement learning DRL
29 Vul-R2: A Reasoning LLM for Automated Vulnerability Repair Vul-R2：一种用于自动漏洞修复的推理LLM reinforcement learning large language model foundation model
30 In-the-Flow Agentic System Optimization for Effective Planning and Tool Use 提出AgentFlow，通过在流程中优化Agent系统，有效提升规划能力和工具使用效果 reinforcement learning large language model
31 Towards Reliable and Practical LLM Security Evaluations via Bayesian Modelling 提出基于贝叶斯建模的LLM安全评估框架，提升prompt注入攻击漏洞评估的可靠性与实用性 Mamba large language model
32 TaTToo: Tool-Grounded Thinking PRM for Test-Time Scaling in Tabular Reasoning 提出TaTToo，一种工具驱动的表格推理PRM，提升测试时表格推理能力。 reinforcement learning reward shaping

🔬 支柱一：机器人控制 (Robot Control) (1 篇)

#题目一句话要点标签🔗⭐
33 D2E: Scaling Vision-Action Pretraining on Desktop Data for Transfer to Embodied AI D2E框架：利用桌面数据预训练提升具身智能机器人性能 manipulation embodied AI 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🔗⭐
34 Mitigating Surgical Data Imbalance with Dual-Prediction Video Diffusion Model 提出SurgiFlowVid，利用双预测视频扩散模型缓解手术视频数据不平衡问题 scene understanding optical flow motion prediction

🔬 支柱四：生成式动作 (Generative Motion) (1 篇)

#题目一句话要点标签🔗⭐
35 AutoPentester: An LLM Agent-based Framework for Automated Pentesting AutoPentester：基于LLM Agent的自动化渗透测试框架 penetration large language model

⬅️ 返回 cs.AI 首页 · 🏠 返回主页