cs.AI（2025-04-29）

📊 共 24 篇论文

🎯 兴趣领域导航

支柱九：具身大模型 (Embodied Foundation Models) (18) 支柱二：RL算法与架构 (RL & Architecture) (4) 支柱五：交互与反应 (Interaction & Reaction) (1) 支柱八：物理动画 (Physics-based Animation) (1)

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

#	题目	一句话要点	标签	🔗	⭐
1	On the Potential of Large Language Models to Solve Semantics-Aware Process Mining Tasks	探索大语言模型在语义感知过程挖掘任务中的潜力	large language model
2	CoCo-Bench: A Comprehensive Code Benchmark For Multi-task Large Language Model Evaluation	CoCo-Bench：用于多任务大语言模型代码能力评估的综合基准	large language model
3	The Hidden Risks of LLM-Generated Web Application Code: A Security-Centric Evaluation of Code Generation Capabilities in Large Language Models	评估LLM生成Web应用代码的安全性风险，揭示潜在漏洞	large language model
4	CachePrune: Neural-Based Attribution Defense Against Indirect Prompt Injection Attacks	CachePrune：基于神经元归因的防御间接提示注入攻击方法	large language model instruction following
5	Using LLMs in Generating Design Rationale for Software Architecture Decisions	利用大型语言模型生成软件架构决策的设计原理	large language model chain-of-thought
6	TAMO: Fine-Grained Root Cause Analysis via Tool-Assisted LLM Agent with Multi-Modality Observation Data in Cloud-Native Systems	TAMO：云原生系统中基于工具辅助LLM Agent的多模态细粒度根因分析	large language model multimodal
7	Theoretical Foundations for Semantic Cognition in Artificial Intelligence	提出基于结构化语义状态的模块化认知架构，用于构建具备反思能力的智能体	large language model
8	Ascendra: Dynamic Request Prioritization for Efficient LLM Serving	Ascendra：面向高效LLM服务的动态请求优先级调度系统	large language model
9	Hallucination by Code Generation LLMs: Taxonomy, Benchmarks, Mitigation, and Challenges	综述代码生成LLM的幻觉问题：分类、评测、缓解与挑战	large language model
10	Cognitive maps are generative programs	提出基于生成程序的认知地图模型，提升具身智能体在结构化环境中的规划效率。	large language model
11	Information Retrieval in the Age of Generative AI: The RGB Model	提出RGB模型，量化生成式AI时代信息检索中的真实性与可靠性问题	large language model
12	MuRAL: A Multi-Resident Ambient Sensor Dataset Annotated with Natural Language for Activities of Daily Living	MuRAL：首个多住户环境传感器自然语言标注数据集，用于日常活动识别	large language model
13	Token-Efficient Prompt Injection Attack: Provoking Cessation in LLM Reasoning via Adaptive Token Compression	提出基于自适应Token压缩的推理中断攻击，高效触发LLM的“停止思考”漏洞	large language model
14	Enhancing News Recommendation with Hierarchical LLM Prompting	PNR-LLM：利用分层LLM提示增强新闻推荐，提升用户偏好捕捉能力。	large language model
15	APG-MOS: Auditory Perception Guided-MOS Predictor for Synthetic Speech	APG-MOS：提出听觉感知引导的MOS预测器，提升合成语音质量评估与人类判断一致性	multimodal
16	RV-Syn: Rational and Verifiable Mathematical Reasoning Data Synthesis based on Structured Function Library	RV-Syn：基于结构化函数库的理性可验证数学推理数据合成方法	large language model
17	CrashFixer: A crash resolution agent for the Linux kernel	CrashFixer：一个基于LLM的Linux内核崩溃修复Agent	large language model
18	Skill Discovery for Software Scripting Automation via Offline Simulations with LLMs	提出基于离线模拟和LLM的软件脚本自动化技能发现框架	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	A Survey on GUI Agents with Foundation Models Enhanced by Reinforcement Learning	综述：基于强化学习增强的具身智能GUI Agent	reinforcement learning policy learning large language model
20	ChestX-Reasoner: Advancing Radiology Foundation Models with Reasoning through Step-by-Step Verification	ChestX-Reasoner：通过逐步验证推理提升放射学基础模型	reinforcement learning large language model foundation model
21	Head-Tail-Aware KL Divergence in Knowledge Distillation for Spiking Neural Networks	提出Head-Tail-Aware KL散度，用于SNN知识蒸馏，提升模型泛化能力。	distillation
22	Domain-Agnostic Scalable AI Safety Ensuring Framework	提出领域无关的AI安全框架以确保高性能与安全性	reinforcement learning PPO

🔬 支柱五：交互与反应 (Interaction & Reaction) (1 篇)

#	题目	一句话要点	标签	🔗	⭐
23	Spatiotemporal Emotional Synchrony in Dyadic Interactions: The Role of Speech Conditions in Facial and Vocal Affective Alignment	研究语音条件下人际互动中情绪同步的时空特性，揭示语音重叠对情感协调的影响	dyadic interaction spatiotemporal multimodal

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

#	题目	一句话要点	标签	🔗	⭐
24	DiffusionRIR: Room Impulse Response Interpolation using Diffusion Models	DiffusionRIR：利用扩散模型进行房间脉冲响应插值	PULSE

⬅️ 返回 cs.AI 首页 · 🏠 返回主页