cs.AI（2025-12-16）

📊 共 23 篇论文 | 🔗 3 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) (6 🔗2) 支柱七：动作重定向 (Motion Retargeting) (1) 支柱四：生成式动作 (Generative Motion) (1) 支柱一：机器人控制 (Robot Control) (1)

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

#题目一句话要点标签🔗⭐
1 HydroGEM: A Self Supervised Zero Shot Hybrid TCN Transformer Foundation Model for Continental Scale Streamflow Quality Control HydroGEM：用于洲际尺度流量质量控制的自监督零样本混合TCN-Transformer基础模型 foundation model zero-shot transfer
2 Intention Chain-of-Thought Prompting with Dynamic Routing for Code Generation 提出RoutingGen框架，通过动态路由和意图链式思考提升代码生成性能。 large language model chain-of-thought
3 Sparsity-Controllable Dynamic Top-p MoE for Large Foundation Model Pre-training 提出DTop-p MoE，实现稀疏度可控的动态Top-p路由，提升大模型预训练效果。 large language model foundation model
4 Massive Editing for Large Language Models Based on Dynamic Weight Generation 提出基于动态权重生成的大语言模型批量知识编辑方法MeG large language model
5 PerfCoder: Large Language Models for Interpretable Code Performance Optimization PerfCoder：基于大语言模型的可解释代码性能优化 large language model
6 Model-First Reasoning LLM Agents: Reducing Hallucinations through Explicit Problem Modeling 提出Model-First Reasoning，通过显式建模减少LLM在复杂规划任务中的幻觉 large language model chain-of-thought
7 Leveraging LLMs for Collaborative Ontology Engineering in Parkinson Disease Monitoring and Alerting 利用大型语言模型进行帕金森病监测和预警的协同本体工程 large language model chain-of-thought
8 OpenDataArena: A Fair and Open Arena for Benchmarking Post-Training Dataset Value OpenDataArena：一个公平开放的平台，用于评估后训练数据集的价值 large language model foundation model
9 Seismology modeling agent: A smart assistant for geophysical researchers 提出基于大语言模型的地震学建模智能助手，降低SPECFEM使用门槛。 large language model ✅
10 PortAgent: LLM-driven Vehicle Dispatching Agent for Port Terminals PortAgent：基于LLM的港口车辆调度智能体，提升跨港口迁移能力 large language model
11 TiCard: Deployable EXPLAIN-only Residual Learning for Cardinality Estimation TiCard：一种可部署的、仅使用EXPLAIN信息的基数估计残差学习框架 foundation model
12 SPARQL-LLM: Real-Time SPARQL Query Generation from Natural Language Questions SPARQL-LLM：一种基于轻量级元数据的实时自然语言到SPARQL查询生成方法 large language model
13 IntentMiner: Intent Inversion Attack via Tool Call Analysis in the Model Context Protocol 提出IntentMiner，通过分析工具调用日志实现用户意图反演攻击。 large language model
14 LAPPI: Interactive Optimization with LLM-Assisted Preference-Based Problem Instantiation LAPPI：利用LLM辅助的偏好问题实例化进行交互式优化 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🔗⭐
15 RADAR: Accelerating Large Language Model Inference With RL-Based Dynamic Draft Trees RADAR：基于强化学习的动态草稿树加速大语言模型推理 reinforcement learning offline reinforcement learning large language model ✅
16 Incentivizing Tool-augmented Thinking with Images for Medical Image Analysis Ophiuchus：一种工具增强的医学图像分析框架，提升MLLM的细粒度推理能力 reinforcement learning multimodal chain-of-thought
17 Context-Picker: Dynamic context selection using multi-stage reinforcement learning Context-Picker：利用多阶段强化学习动态选择长文本问答上下文 reinforcement learning distillation reward shaping
18 Evaluating Small Language Models for Agentic On-Farm Decision Support Systems 评估小型语言模型在农场决策支持系统中的应用潜力，Qwen-4B表现突出。 predictive model large language model
19 MobileWorldBench: Towards Semantic World Modeling For Mobile Agents 提出MobileWorldBench，利用视觉-语言模型为移动Agent构建语义世界模型 world model ✅
20 A Threshold-Triggered Deep Q-Network-Based Framework for Self-Healing in Autonomic Software-Defined IIoT-Edge Networks 提出基于阈值触发深度Q网络的自愈框架，用于软件定义IIoT边缘网络 reinforcement learning deep reinforcement learning

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

#题目一句话要点标签🔗⭐
21 Georeferencing complex relative locality descriptions with large language models 利用大型语言模型解决生物多样性领域复杂相对位置描述的地理定位问题 spatial relationship large language model

🔬 支柱四：生成式动作 (Generative Motion) (1 篇)

#题目一句话要点标签🔗⭐
22 PentestEval: Benchmarking LLM-based Penetration Testing with Modular and Stage-Level Design PentestEval：首个模块化、分阶段评估LLM渗透测试能力的综合基准 penetration large language model

🔬 支柱一：机器人控制 (Robot Control) (1 篇)

#题目一句话要点标签🔗⭐
23 A data-physics hybrid generative model for patient-specific post-stroke motor rehabilitation using wearable sensor data 提出数据-物理混合生成模型，利用可穿戴传感器数据实现卒中后患者的个性化运动康复。 locomotion reinforcement learning deep reinforcement learning

⬅️ 返回 cs.AI 首页 · 🏠 返回主页