cs.AI（2026-03-16）

📊 共 24 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) (14 🔗3) 支柱一：机器人控制 (Robot Control) (4) 支柱二：RL算法与架构 (RL & Architecture) (4) 支柱七：动作重定向 (Motion Retargeting) (1 🔗1) 支柱四：生成式动作 (Generative Motion) (1)

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

#题目一句话要点标签🔗⭐
1 Advancing Multimodal Agent Reasoning with Long-Term Neuro-Symbolic Memory 提出NS-Mem神经符号记忆框架，提升多模态Agent在复杂环境下的推理能力。 large language model multimodal
2 VTC-Bench: Evaluating Agentic Multimodal Models via Compositional Visual Tool Chaining VTC-Bench：通过组合式视觉工具链评估Agentic多模态模型 large language model multimodal
3 BrainBench: Exposing the Commonsense Reasoning Gap in Large Language Models BrainBench：揭示大型语言模型中常识推理的差距 large language model
4 Brain-Inspired Graph Multi-Agent Systems for LLM Reasoning 提出脑启发图多智能体系统BIGMAS，提升LLM复杂推理能力 large language model chain-of-thought
5 OpenSeeker: Democratizing Frontier Search Agents by Fully Open-Sourcing Training Data OpenSeeker：通过完全开源训练数据，实现前沿搜索Agent的普及化。 large language model
6 InterveneBench: Benchmarking LLMs for Intervention Reasoning and Causal Study Design in Real Social Systems InterveneBench：评估LLM在真实社会系统中干预推理和因果研究设计能力 large language model ✅
7 Unlocking the Value of Text: Event-Driven Reasoning and Multi-Level Alignment for Time Series Forecasting 提出VoT，利用事件驱动推理和多层次对齐，提升文本增强时间序列预测性能。 multimodal ✅
8 SKILLS: Structured Knowledge Injection for LLM-Driven Telecommunications Operations SKILLS：通过结构化知识注入提升LLM在电信运营中的可靠性 large language model
9 PMAx: An Agentic Framework for AI-Driven Process Mining PMAx：一个用于AI驱动的过程挖掘的Agentic框架，解决LLM直接应用于过程挖掘的局限性。 large language model
10 Why the Valuable Capabilities of LLMs Are Precisely the Unexplainable Ones 论证大语言模型最有价值的能力恰恰是那些无法解释的部分 large language model
11 To See is Not to Master: Teaching LLMs to Use Private Libraries for Code Generation PriCoder：通过数据合成提升LLM在私有库API代码生成中的能力 large language model ✅
12 Why Agents Compromise Safety Under Pressure 揭示Agentic Pressure：压力下大语言模型Agent的安全妥协现象 large language model
13 $p^2$RAG: Privacy-Preserving RAG Service Supporting Arbitrary Top-$k$ Retrieval 提出$p^2$RAG，一种支持任意Top-$k$检索的隐私保护RAG服务。 large language model
14 Beyond Local Code Optimization: Multi-Agent Reasoning for Software System Optimization 提出基于多智能体推理的软件系统优化框架，提升微服务性能。 large language model

🔬 支柱一：机器人控制 (Robot Control) (4 篇)

#题目一句话要点标签🔗⭐
15 SFCoT: Safer Chain-of-Thought via Active Safety Evaluation and Calibration 提出SFCoT框架，通过主动安全评估与校准增强LLM推理过程的安全性 manipulation large language model chain-of-thought
16 InterPol: De-anonymizing LM Arena via Interpolated Preference Learning 提出INTERPOL，通过插值偏好学习破解LM Arena的模型匿名性。 manipulation preference learning curriculum learning
17 Are Dilemmas and Conflicts in LLM Alignment Solvable? A View from Priority Graph 构建优先级图模型，分析LLM对齐困境并提出运行时验证机制 manipulation large language model
18 SCAN: Sparse Circuit Anchor Interpretable Neuron for Lifelong Knowledge Editing 提出SCAN：基于稀疏电路锚定神经元的终身知识编辑框架，解决LLM灾难性遗忘问题 manipulation 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🔗⭐
19 RS-WorldModel: a Unified Model for Remote Sensing Understanding and Future Sense Forecasting 提出RS-WorldModel，统一遥感理解与未来场景预测，性能超越更大规模模型。 world model spatiotemporal
20 Listening to the Echo: User-Reaction Aware Policy Optimization via Scalar-Verbal Hybrid Reinforcement Learning 提出RAPO框架，通过标量-文本混合强化学习优化用户反应驱动的情感支持对话策略。 reinforcement learning distillation
21 SAGE: Multi-Agent Self-Evolution for LLM Reasoning SAGE：面向LLM推理的多智能体自进化框架，提升数学和代码生成能力 reinforcement learning large language model
22 Interference-Aware K-Step Reachable Communication in Multi-Agent Reinforcement Learning 提出IA-KRC框架，解决多智能体强化学习中干扰感知通信问题 reinforcement learning

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

#题目一句话要点标签🔗⭐
23 Intelligent Co-Design: An Interactive LLM Framework for Interior Spatial Design via Multi-Modal Agents 提出基于LLM的多模态交互式室内空间智能协同设计框架，提升设计效率与用户参与度。 spatial relationship large language model multimodal ✅

🔬 支柱四：生成式动作 (Generative Motion) (1 篇)

#题目一句话要点标签🔗⭐
24 Architecture-Agnostic Feature Synergy for Universal Defense Against Heterogeneous Generative Threats 提出架构无关的特征协同框架ATFS，实现对异构生成威胁的通用防御 VQ-VAE

⬅️ 返回 cs.AI 首页 · 🏠 返回主页