cs.AI（2025-02-14）

📊 共 26 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) (14 🔗1) 支柱二：RL算法与架构 (RL & Architecture) (10) 支柱一：机器人控制 (Robot Control) (1) 支柱八：物理动画 (Physics-based Animation) (1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (14 篇)

#题目一句话要点标签🔗⭐
1 Do Large Language Models Reason Causally Like Us? Even Better? 评估大语言模型因果推理能力：部分模型超越人类，但仍有局限 large language model
2 Decision Information Meets Large Language Models: The Future of Explainable Operations Research 提出可解释的运筹学框架以解决决策透明性问题 large language model
3 Has My System Prompt Been Used? Large Language Model Prompt Membership Inference Prompt Detective：基于输出分布差异的大语言模型提示词成员推断方法 large language model
4 MuDoC: An Interactive Multimodal Document-grounded Conversational AI System 提出MuDoC：一个交互式多模态文档对话AI系统，支持图文混合的文档内容理解与交互。 multimodal
5 Automatic Evaluation Metrics for Artificially Generated Scientific Research 提出基于引用预测和评审评分预测的自动评估指标，用于评估AI生成的科学研究 large language model foundation model
6 Artificial Intelligence in Spectroscopy: Advancing Chemistry from Prediction to Generation and Beyond 综述光谱机器学习：从预测到生成，推进化学领域AI应用 foundation model multimodal ✅
7 AutoS$^2$earch: Unlocking the Reasoning Potential of Large Models for Web-based Source Search AutoS$^2$earch：利用大模型进行Web环境下的零样本源搜索 large language model chain-of-thought
8 MEADOW: Memory-efficient Dataflow and Data Packing for Low Power Edge LLMs MEADOW：面向低功耗边缘LLM的内存高效数据流和数据打包 large language model
9 GraphiT: Efficient Node Classification on Text-Attributed Graphs with Prompt Optimized LLMs GraphiT：利用提示优化LLM实现文本属性图上的高效节点分类 large language model
10 LLM-Powered Preference Elicitation in Combinatorial Assignment 提出基于LLM的偏好启发框架，提升组合分配中的资源分配效率 large language model
11 MIR-Bench: Can Your LLM Recognize Complicated Patterns via Many-Shot In-Context Reasoning? MIR-Bench：提出多示例上下文推理基准，评估LLM在复杂模式识别中的能力 large language model
12 MathConstruct: Challenging LLM Reasoning with Constructive Proofs MathConstruct：提出构造性证明数学基准，挑战LLM推理能力 large language model
13 The Ann Arbor Architecture for Agent-Oriented Programming 提出Ann Arbor架构，用于面向Agent的大语言模型编程，优化上下文学习。 large language model
14 ArchRAG: Attributed Community-based Hierarchical Retrieval-Augmented Generation 提出ArchRAG，利用属性社区分层检索增强生成，提升图数据问答准确率并降低token成本。 large language model

🔬 支柱二：RL算法与架构 (RL & Architecture) (10 篇)

#题目一句话要点标签🔗⭐
15 A Self-Supervised Reinforcement Learning Approach for Fine-Tuning Large Language Models Using Cross-Attention Signals 提出一种基于自监督强化学习的LLM微调方法，利用交叉注意力信号作为奖励。 reinforcement learning RLHF large language model
16 POI-Enhancer: An LLM-based Semantic Enhancement Framework for POI Representation Learning POI-Enhancer：基于LLM的POI表示学习语义增强框架 representation learning contrastive learning large language model
17 ClusMFL: A Cluster-Enhanced Framework for Modality-Incomplete Multimodal Federated Learning in Brain Imaging Analysis 提出ClusMFL框架，解决脑影像分析中模态不完全多模态联邦学习问题。 contrastive learning multimodal
18 Hybrid Offline-online Scheduling Method for Large Language Model Inference Optimization 提出混合离线-在线调度方法，优化大语言模型推理服务系统吞吐量。 reinforcement learning large language model
19 Unlocking Efficient Vehicle Dynamics Modeling via Analytic World Models 提出基于可微世界模型的车辆动力学高效建模方法，用于自动驾驶等场景。 world model differentiable simulation
20 Towards Empowerment Gain through Causal Structure Learning in Model-Based RL 提出ECL框架，通过因果结构学习提升模型强化学习中的控制能力和样本效率。 reinforcement learning model-based RL
21 Tempo: Helping Data Scientists and Domain Experts Collaboratively Specify Predictive Modeling Tasks Tempo：协同式预测建模任务规范交互系统，提升数据科学家与领域专家合作效率 predictive model
22 Reinforcement Learning in Strategy-Based and Atari Games: A Review of Google DeepMinds Innovations 综述DeepMind在策略游戏与Atari游戏中基于强化学习的创新 reinforcement learning
23 Cooperative Multi-Agent Planning with Adaptive Skill Synthesis COMPASS：基于自适应技能合成的合作式多智能体规划框架 reinforcement learning large language model
24 Diverse Inference and Verification for Advanced Reasoning 提出多样化推理与验证方法，显著提升LLM在复杂推理任务上的性能 reinforcement learning IMoS

🔬 支柱一：机器人控制 (Robot Control) (1 篇)

#题目一句话要点标签🔗⭐
25 Causal Information Prioritization for Efficient Reinforcement Learning 提出因果信息优先级（CIP）方法，提升强化学习在复杂环境中的样本效率。 locomotion manipulation reinforcement learning

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

#题目一句话要点标签🔗⭐
26 Machine Learning for Phase Estimation in Satellite-to-Earth Quantum Communication 提出低复杂度LSTM神经网络，用于卫星-地球量子通信中的相位估计，提升CV-QKD实时性能。 PULSE

⬅️ 返回 cs.AI 首页 · 🏠 返回主页