cs.AI（2025-03-14）

📊 共 20 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) (15 🔗2) 支柱二：RL算法与架构 (RL & Architecture) (5 🔗1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (15 篇)

#题目一句话要点标签🔗⭐
1 Conversational AI as a Coding Assistant: Understanding Programmers' Interactions with and Expectations from Large Language Models for Coding 研究程序员与LLM代码助手交互模式，提出改进设计指南 large language model multimodal
2 Synthesizing Access Control Policies using Large Language Models 利用大型语言模型合成访问控制策略，提升云环境安全性。 large language model
3 Potential of large language model-powered nudges for promoting daily water and energy conservation 利用大语言模型赋能的“助推”手段，促进日常节水节能行为 large language model
4 Large language model-powered AI systems achieve self-replication with no human intervention 大型语言模型驱动的AI系统在无人干预下实现自我复制 large language model
5 Optimizing Large Language Models for Detecting Symptoms of Comorbid Depression or Anxiety in Chronic Diseases: Insights from Patient Messages 优化大型语言模型以检测慢性病患者的共病抑郁或焦虑症状 large language model
6 Visualizing Thought: Conceptual Diagrams Enable Robust Planning in LMMs 提出Visual Thinking框架，利用概念图提升LMM在复杂规划任务中的性能 large language model multimodal
7 ASMA-Tune: Unlocking LLMs' Assembly Code Comprehension via Structural-Semantic Instruction Tuning ASMA-Tune：通过结构-语义指令调优解锁大语言模型对汇编代码的理解能力 large language model instruction following
8 API Agents vs. GUI Agents: Divergence and Convergence 首个API与GUI型LLM智能体对比研究，揭示其差异、融合与应用场景 large language model multimodal
9 Safety Mirage: How Spurious Correlations Undermine VLM Safety Fine-Tuning and Can Be Mitigated by Machine Unlearning 揭示视觉语言模型安全微调中的“安全幻影”现象，并提出基于机器遗忘的缓解方案 multimodal
10 Can Large Reasoning Models do Analogical Reasoning under Perceptual Uncertainty? 评估大型推理模型在感知不确定性下的类比推理能力，发现其鲁棒性不足 large language model ✅
11 Integrating LLMs in Gamified Systems 提出LLM集成框架，提升游戏化系统任务动态性与用户参与度 large language model
12 Cerebrum (AIOS SDK): A Platform for Agent Development, Deployment, Distribution, and Discovery Cerebrum：AIOS SDK，用于自主Agent的开发、部署、分发与发现 chain-of-thought ✅
13 LLM Agents for Education: Advances and Applications LLM Agent在教育领域应用综述：自动化教学任务与领域定制Agent large language model
14 Cross-Modal Learning for Music-to-Music-Video Description Generation 提出音乐到音乐视频描述生成方法，利用跨模态学习弥合音乐与视频的差异。 multimodal
15 Exploring the Necessity of Reasoning in LLM-based Agent Scenarios LaRMA框架揭示LLM与LRM在Agent任务中的推理必要性与权衡 large language model

🔬 支柱二：RL算法与架构 (RL & Architecture) (5 篇)

#题目一句话要点标签🔗⭐
16 Policy Frameworks for Transparent Chain-of-Thought Reasoning in Large Language Models 提出分层访问策略框架，平衡LLM中CoT推理的透明度、安全性和责任 distillation large language model chain-of-thought
17 Reinforcement Learning Outperforms Supervised Fine-Tuning: A Case Study on Audio Question Answering 强化学习在音频问答任务中超越监督微调，实现更优性能 reinforcement learning large language model multimodal ✅
18 Combinatorial Optimization for All: Using LLMs to Aid Non-Experts in Improving Optimization Algorithms 利用大型语言模型辅助非专家改进组合优化算法 reinforcement learning large language model
19 Monitoring Reasoning Models for Misbehavior and the Risks of Promoting Obfuscation 利用CoT监控检测并抑制AI奖励攻击，但需警惕过度优化导致的意图混淆。 reinforcement learning chain-of-thought
20 GKG-LLM: A Unified Framework for Generalized Knowledge Graph Construction 提出GKG-LLM统一框架，用于广义知识图谱构建，提升跨领域和任务性能。 curriculum learning large language model

⬅️ 返回 cs.AI 首页 · 🏠 返回主页