cs.AI（2025-02-13）

📊 共 31 篇论文 | 🔗 2 篇有代码

🎯 兴趣领域导航

支柱九：具身大模型 (Embodied Foundation Models) (21 🔗2) 支柱一：机器人控制 (Robot Control) (3) 支柱二：RL算法与架构 (RL & Architecture) (3) 支柱三：空间感知与语义 (Perception & Semantics) (2) 支柱四：生成式动作 (Generative Motion) (1) 支柱五：交互与反应 (Interaction & Reaction) (1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (21 篇)

#题目一句话要点标签🔗⭐
1 From large language models to multimodal AI: A scoping review on the potential of generative AI in medicine 综述性研究：大型语言模型到多模态AI在医学领域的应用与潜力 large language model multimodal
2 MDCrow: Automating Molecular Dynamics Workflows with Large Language Models MDCrow：利用大型语言模型自动化分子动力学工作流程 large language model chain-of-thought
3 DreamLLM-3D: Affective Dream Reliving using Large Language Model and 3D Generative AI DreamLLM-3D：利用大语言模型和3D生成AI实现情感化的梦境重现 large language model multimodal
4 Game Theory Meets Large Language Models: A Systematic Survey with Taxonomy and New Frontiers 首次全面综述博弈论与大语言模型的双向关系，并提出新的分类框架。 large language model
5 Brain-Inspired Exploration of Functional Networks and Key Neurons in Large Language Models 受脑启发，探索大语言模型中的功能网络与关键神经元 large language model ✅
6 QueryAttack: Jailbreaking Aligned Large Language Models Using Structured Non-natural Query Language 提出QueryAttack以破解大型语言模型的安全防护 large language model ✅
7 CoT-Valve: Length-Compressible Chain-of-Thought Tuning 提出CoT-Valve，通过可控的思维链长度调整推理模型，降低推理成本。 chain-of-thought
8 EnigmaEval: A Benchmark of Long Multimodal Reasoning Challenges 提出EnigmaEval：一个长程多模态推理挑战基准，用于评估语言模型的认知能力。 multimodal
9 Visual Graph Question Answering with ASP and LLMs for Language Parsing 提出结合ASP和LLM的VQA方法，解决图结构图像的问答任务 large language model multimodal
10 AgentGuard: Repurposing Agentic Orchestrator for Safety Evaluation of Tool Orchestration AgentGuard：利用Agent编排器进行工具编排安全评估 large language model
11 Toward Total Recall: Enhancing FAIRness through AI-Driven Metadata Standardization 利用AI驱动的元数据标准化提升数据检索的全面性 large language model
12 TableTalk: Scaffolding Spreadsheet Development with a Language Agent TableTalk：利用语言Agent脚手架式地辅助电子表格开发 large language model
13 KIMAs: A Configurable Knowledge Integrated Multi-Agent System KIMAs：一个可配置的知识集成多智能体系统，用于构建知识密集型应用。 large language model
14 On LLM-generated Logic Programs and their Inference Execution Methods 利用LLM生成逻辑程序，并探索其推理执行方法，提升LLM知识利用率。 large language model
15 Logical Lease Litigation: Prolog and LLMs for Rental Law Compliance in New York LogicLease：结合Prolog和LLM，实现纽约州租赁法合规自动化分析 large language model
16 FLAME: Flexible LLM-Assisted Moderation Engine FLAME：一种灵活的LLM辅助内容审核引擎，有效防御对抗性攻击。 large language model
17 Show Me the Work: Fact-Checkers' Requirements for Explainable Automated Fact-Checking 通过访谈分析，揭示事实核查人员对可解释自动化事实核查工具的需求。 large language model
18 Self-Consistency of the Internal Reward Models Improves Self-Rewarding Language Models 提出SCIR框架，提升自奖励语言模型内部奖励模型的一致性，从而提高对齐性能。 large language model
19 MIH-TCCT: Mitigating Inconsistent Hallucinations in LLMs via Event-Driven Text-Code Cyclic Training 提出MIH-TCCT框架，通过事件驱动的文本-代码循环训练缓解LLM中的不一致性幻觉问题 large language model
20 Learning in Strategic Queuing Systems with Small Buffers 在具有小缓冲区的策略性排队系统中实现学习，提升系统稳定性 TAMP
21 Application Modernization with LLMs: Addressing Core Challenges in Reliability, Security, and Quality 提出融合代码推理与生成的大语言模型框架，提升应用现代化改造的可靠性与安全性 large language model

🔬 支柱一：机器人控制 (Robot Control) (3 篇)

#题目一句话要点标签🔗⭐
22 Diverse Transformer Decoding for Offline Reinforcement Learning Using Financial Algorithmic Approaches 提出Portfolio Beam Search，提升离线强化学习Transformer解码的多样性和鲁棒性 locomotion reinforcement learning offline RL
23 EmbodiedBench: Comprehensive Benchmarking Multi-modal Large Language Models for Vision-Driven Embodied Agents EmbodiedBench：用于视觉驱动具身智能体的多模态大语言模型综合评测基准 manipulation large language model
24 AIvaluateXR: An Evaluation Framework for on-Device AI in XR with Benchmarking Results AIvaluateXR：用于XR设备端AI的评估框架与基准测试 Apple Vision Pro large language model

🔬 支柱二：RL算法与架构 (RL & Architecture) (3 篇)

#题目一句话要点标签🔗⭐
25 Logical Reasoning in Large Language Models: A Survey 综述大型语言模型中的逻辑推理能力，并分析提升策略与未来方向 reinforcement learning large language model
26 Reinforced Large Language Model is a formal theorem prover 提出基于强化学习的大语言模型定理证明框架，提升证明准确率 reinforcement learning large language model
27 MC2SleepNet: Multi-modal Cross-masking with Contrastive Learning for Sleep Stage Classification MC2SleepNet：基于多模态跨掩码和对比学习的睡眠分期网络 contrastive learning

🔬 支柱三：空间感知与语义 (Perception & Semantics) (2 篇)

#题目一句话要点标签🔗⭐
28 X-SG$^2$S: Safe and Generalizable Gaussian Splatting with X-dimensional Watermarks X-SG$^2$S：通过X维水印实现安全且可泛化的高斯溅射 3D gaussian splatting 3DGS gaussian splatting
29 Co-designing Large Language Model Tools for Project-Based Learning with K12 Educators 通过与K12教师共创，设计基于大语言模型的项目制学习工具 affordance large language model

🔬 支柱四：生成式动作 (Generative Motion) (1 篇)

#题目一句话要点标签🔗⭐
30 PenTest++: Elevating Ethical Hacking with AI and Automation PenTest++：利用AI和自动化提升伦理黑客效率与可扩展性 penetration

🔬 支柱五：交互与反应 (Interaction & Reaction) (1 篇)

#题目一句话要点标签🔗⭐
31 Setup Once, Secure Always: A Single-Setup Secure Federated Learning Aggregation Protocol with Forward and Backward Secrecy for Dynamic Users 提出一种支持动态用户、具备前后向安全性的单次设置联邦学习安全聚合协议 OMOMO

⬅️ 返回 cs.AI 首页 · 🏠 返回主页