cs.AI（2025-04-14）

📊 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 🔗2) 支柱二：RL算法与架构 (RL & Architecture) (4 🔗1) 支柱一：机器人控制 (Robot Control) (2 🔗1) 支柱四：生成式动作 (Generative Motion) (1) 支柱七：动作重定向 (Motion Retargeting) (1) 支柱八：物理动画 (Physics-based Animation) (1)

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

#题目一句话要点标签🔗⭐
1 The Future of MLLM Prompting is Adaptive: A Comprehensive Experimental Evaluation of Prompt Engineering Methods for Robust Multimodal Performance 多模态大语言模型提示工程自适应优化：全面实验评估提升模型鲁棒性 large language model multimodal chain-of-thought
2 See or Recall: A Sanity Check for the Role of Vision in Solving Visualization Question Answer Tasks with Multimodal LLMs 提出VisQA数据集的健全性检查框架，区分多模态LLM的视觉推理与知识回忆能力 large language model multimodal
3 MatterTune: An Integrated, User-Friendly Platform for Fine-Tuning Atomistic Foundation Models to Accelerate Materials Simulation and Discovery MatterTune：集成化平台，微调原子级预训练模型，加速材料模拟与发现 foundation model
4 Building Trustworthy Multimodal AI: A Review of Fairness, Transparency, and Ethics in Vision-Language Tasks 综述：构建可信赖的多模态AI，关注视觉-语言任务中的公平性、透明性和伦理 multimodal
5 Optimizing Data Distribution and Kernel Performance for Efficient Training of Chemistry Foundation Models: A Case Study with MACE 优化数据分布与内核性能，加速化学基础模型MACE的训练。 foundation model
6 Can Competition Enhance the Proficiency of Agents Powered by Large Language Models in the Realm of News-driven Time Series Forecasting? 提出基于竞争机制的多智能体新闻驱动时序预测框架，提升LLM在金融预测中的创新能力。 large language model
7 Investigating cybersecurity incidents using large language models in latest-generation wireless networks 利用大型语言模型检测最新一代无线网络中的网络安全事件 large language model
8 Working with Large Language Models to Enhance Messaging Effectiveness for Vaccine Confidence 利用大型语言模型增强疫苗信任信息的有效性 large language model
9 A Survey of Large Language Model-Powered Spatial Intelligence Across Scales: Advances in Embodied Agents, Smart Cities, and Earth Science 综述：大语言模型驱动的跨尺度空间智能研究进展 large language model
10 StruPhantom: Evolutionary Injection Attacks on Black-Box Tabular Agents Powered by Large Language Models StruPhantom：针对黑盒表格代理的进化注入攻击，利用大语言模型 large language model
11 Zero-shot Autonomous Microscopy for Scalable and Intelligent Characterization of 2D Materials ATOMIC：利用基础模型实现二维材料的零样本自主显微表征 large language model foundation model
12 RealWebAssist: A Benchmark for Long-Horizon Web Assistance with Real-World Users RealWebAssist：一个面向真实用户的长时程Web辅助基准测试 instruction following
13 Hierarchical Knowledge Graphs for Story Understanding in Visual Narratives 提出层级知识图谱框架，用于视觉叙事中的故事结构化语义理解 multimodal
14 Breaking the Data Barrier -- Building GUI Agents Through Task Generalization 通过任务泛化构建GUI智能体，突破数据壁垒 multimodal ✅
15 Can LLMs Assist Expert Elicitation for Probabilistic Causal Modeling? 利用LLM辅助概率因果建模专家知识获取，提升生物识别与医疗决策透明度。 large language model
16 SymRTLO: Enhancing RTL Code Optimization with LLMs and Neuron-Inspired Symbolic Reasoning SymRTLO：利用LLM和神经符号推理增强RTL代码优化 large language model
17 Understanding and Optimizing Multi-Stage AI Inference Pipelines HERMES：用于理解和优化多阶段AI推理流水线的异构执行模拟器 large language model
18 AlayaDB: The Data Foundation for Efficient and Effective Long-context LLM Inference AlayaDB：用于高效长文本LLM推理的数据底座 large language model
19 The Code Barrier: What LLMs Actually Understand? 通过代码混淆评估LLM代码理解能力，揭示通用模型意外的鲁棒性 foundation model
20 LLM-Driven NPCs: Cross-Platform Dialogue System for Games and Social Platforms 提出基于LLM的跨平台NPC对话系统，实现游戏与社交平台互动 large language model
21 Privacy Meets Explainability: Managing Confidential Data and Transparency Policies in LLM-Empowered Science DataShield：面向LLM赋能科研的数据泄露检测与隐私策略管理框架 large language model
22 Automated Validation of COBOL to Java Transformation 提出基于符号执行的自动化验证框架，用于COBOL到Java代码转换的正确性验证。 large language model
23 Automated Testing of COBOL to Java Transformation 提出COBOL到Java转换的自动化测试框架，解决LLM转换代码的验证难题 large language model
24 EthosGPT: Mapping Human Value Diversity to Advance Sustainable Development Goals (SDGs) EthosGPT：通过映射人类价值观多样性促进可持续发展目标 large language model
25 PestMA: LLM-based Multi-Agent System for Informed Pest Management PestMA：基于LLM的多智能体系统，用于智能病虫害管理 large language model
26 Two Heads are Better Than One: Test-time Scaling of Multi-agent Collaborative Reasoning 提出自适应多智能体框架，通过模型训练和系统协调增强协同推理能力 large language model ✅

🔬 支柱二：RL算法与架构 (RL & Architecture) (4 篇)

#题目一句话要点标签🔗⭐
27 OVERLORD: Ultimate Scaling of DataLoader for Multi-Source Large Foundation Model Training OVERLORD：面向多源大模型训练的数据加载器极致扩展方案 curriculum learning foundation model multimodal
28 Pay Attention to What and Where? Interpretable Feature Extractor in Vision-based Deep Reinforcement Learning 提出可解释特征提取器IFE，解决视觉深度强化学习中注意力漂移问题 reinforcement learning deep reinforcement learning
29 Heimdall: test-time scaling on the generative verification 提出Heimdall，通过生成式验证提升LLM在复杂问题上的推理和验证能力 reinforcement learning chain-of-thought
30 Offline Dynamic Inventory and Pricing Strategy: Addressing Censored and Dependent Demand 提出离线动态库存和定价策略，解决审查和依赖需求下的优化问题 reinforcement learning offline reinforcement learning ✅

🔬 支柱一：机器人控制 (Robot Control) (2 篇)

#题目一句话要点标签🔗⭐
31 Controllable Expressive 3D Facial Animation via Diffusion in a Unified Multimodal Space 提出基于扩散模型的统一多模态空间，实现可控的富有表现力的3D面部动画 manipulation contrastive learning multimodal ✅
32 Do We Really Need Curated Malicious Data for Safety Alignment in Multi-modal Large Language Models? 通过修正数据偏见，无需恶意数据即可提升多模态大语言模型的安全性对齐。 manipulation large language model

🔬 支柱四：生成式动作 (Generative Motion) (1 篇)

#题目一句话要点标签🔗⭐
33 Benchmarking Practices in LLM-driven Offensive Security: Testbeds, Metrics, and Experiment Design 分析LLM驱动的渗透测试基准实践，为未来研究提供行动建议 penetration large language model

🔬 支柱七：动作重定向 (Motion Retargeting) (1 篇)

#题目一句话要点标签🔗⭐
34 Towards Accurate Forecasting of Renewable Energy : Building Datasets and Benchmarking Machine Learning Models for Solar and Wind Power in France 构建法国太阳能和风能预测数据集，并基准测试机器学习模型 spatial relationship

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

#题目一句话要点标签🔗⭐
35 Ride-pool Assignment Algorithms: Modern Implementation and Swapping Heuristics 提出基于交换启发式的拼车分配算法，优化服务率与计算效率，并开源C++实现。 spatiotemporal

⬅️ 返回 cs.AI 首页 · 🏠 返回主页