cs.AI（2025-07-28）

📊 共 26 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) (18 🔗3) 支柱二：RL算法与架构 (RL & Architecture) (6) 支柱八：物理动画 (Physics-based Animation) (1) 支柱一：机器人控制 (Robot Control) (1 🔗1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (18 篇)

#	题目	一句话要点	标签	🔗	⭐
1	An analysis of AI Decision under Risk: Prospect theory emerges in Large Language Models	首次验证：大型语言模型在风险决策中表现出前景理论偏差	large language model chain-of-thought
2	Aligning Large Language Model Agents with Rational and Moral Preferences: A Supervised Fine-Tuning Approach	通过监督微调对齐大语言模型智能体与理性和道德偏好	large language model
3	Enhancing Large Multimodal Models with Adaptive Sparsity and KV Cache Compression	提出自适应稀疏化与KV缓存压缩方法，提升大模型在边缘设备上的部署效率。	multimodal
4	Pareto-Grid-Guided Large Language Models for Fast and High-Quality Heuristics Design in Multi-Objective Combinatorial Optimization	提出基于Pareto网格引导的大语言模型进化算法，用于快速高质量的多目标组合优化启发式设计。	large language model	✅
5	MMGraphRAG: Bridging Vision and Language with Interpretable Multimodal Knowledge Graphs	提出MMGraphRAG，利用多模态知识图谱增强视觉语言检索增强生成任务	multimodal
6	How Chain-of-Thought Works? Tracing Information Flow from Decoding, Projection, and Activation	通过追踪信息流，揭示思维链（CoT）提示的工作机制	chain-of-thought
7	Agentic Web: Weaving the Next Web with AI Agents	构建Agentic Web：利用AI Agent实现自主、目标驱动的互联网交互	large language model	✅
8	Prescriptive Agents based on RAG for Automated Maintenance (PARAM)	PARAM：基于RAG的工业设备预测性维护智能体，实现故障诊断与维护建议自动化。	large language model
9	MIMII-Agent: Leveraging LLMs with Function Calling for Relative Evaluation of Anomalous Sound Detection	提出MIMII-Agent以解决无真实异常声数据的评估问题	large language model
10	Teaching Language Models To Gather Information Proactively	提出主动信息收集框架，提升LLM在复杂任务中作为协作伙伴的能力。	large language model
11	MAAD: Automate Software Architecture Design through Knowledge-Driven Multi-Agent Collaboration	MAAD：通过知识驱动的多智能体协作实现软件架构设计的自动化	large language model
12	Curiosity by Design: An LLM-based Coding Assistant Asking Clarification Questions	设计好奇心：基于LLM的编码助手通过提问进行澄清	large language model
13	LeMix: Unified Scheduling for LLM Training and Inference on Multi-GPU Systems	LeMix：多GPU系统上LLM训练与推理的统一调度系统	large language model
14	CompoST: A Benchmark for Analyzing the Ability of LLMs To Compositionally Interpret Questions in a QALD Setting	CompoST：评估LLM在QALD环境中组合性理解问题的基准测试	large language model
15	A Survey of Self-Evolving Agents: What, When, How, and Where to Evolve on the Path to Artificial Super Intelligence	首个自进化Agent综述：系统性地研究了通向通用人工智能的自进化Agent的设计要素与未来方向。	large language model
16	MIRAGE-Bench: LLM Agent is Hallucinating and Where to Find Them	MIRAGE-Bench：首个交互式LLM Agent幻觉行为统一评测基准	large language model
17	TypyBench: Evaluating LLM Type Inference for Untyped Python Repositories	TypyBench：评估LLM在无类型Python仓库中的类型推断能力	large language model	✅
18	The Xeno Sutra: Can Meaning and Value be Ascribed to an AI-Generated "Sacred" Text?	利用大型语言模型生成佛教“经文”，探讨AI在意义创造领域的哲学和社会影响	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	🔗	⭐
19	Deep Reinforcement Learning-based Cell DTX/DRX Configuration for Network Energy Saving	提出基于深度强化学习的Cell DTX/DRX配置方法，实现5G网络节能。	reinforcement learning deep reinforcement learning DRL
20	Complementarity-driven Representation Learning for Multi-modal Knowledge Graph Completion	提出MoCME框架，利用互补性学习提升多模态知识图谱补全效果	representation learning multimodal
21	Handoff Design in User-Centric Cell-Free Massive MIMO Networks Using DRL	提出基于DRL的用户中心Cell-Free Massive MIMO网络切换设计方案	reinforcement learning deep reinforcement learning DRL
22	Why Flow Matching is Particle Swarm Optimization?	揭示流匹配与粒子群优化对偶性，为生成模型与进化计算融合奠定基础	flow matching
23	JAM: A Tiny Flow-based Song Generator with Fine-grained Controllability and Aesthetic Alignment	JAM：一种基于Flow的小型歌曲生成器，具备细粒度可控性和美学对齐能力	flow matching direct preference optimization
24	Unlearning of Knowledge Graph Embedding via Preference Optimization	提出GraphDPO，通过偏好优化实现知识图谱嵌入的有效不可学习，提升知识完整性。	DPO direct preference optimization

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

#	题目	一句话要点	标签	🔗	⭐
25	Implicit Spatiotemporal Bandwidth Enhancement Filter by Sine-activated Deep Learning Model for Fast 3D Photoacoustic Tomography	提出基于正弦激活深度学习模型的时空带宽增强滤波器，用于快速3D光声层析成像。	spatiotemporal

🔬 支柱一：机器人控制 (Robot Control) (1 篇)

#	题目	一句话要点	标签	🔗	⭐
26	Enhancing Jailbreak Attacks on LLMs via Persona Prompts	提出基于遗传算法的Persona Prompt方法，提升LLM越狱攻击效果	manipulation large language model	✅

⬅️ 返回 cs.AI 首页 · 🏠 返回主页