cs.AI（2025-08-14）

📊 共 23 篇论文 | 🔗 2 篇有代码

🎯 兴趣领域导航

支柱九：具身大模型 (Embodied Foundation Models) (15 🔗1) 支柱二：RL算法与架构 (RL & Architecture) (5 🔗1) 支柱一：机器人控制 (Robot Control) (2) 支柱八：物理动画 (Physics-based Animation) (1)

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

#	题目	一句话要点	标签	🔗	⭐
1	MM-Food-100K: A 100,000-Sample Multimodal Food Intelligence Dataset with Verifiable Provenance	提出MM-Food-100K多模态食物数据集，用于提升图像营养预测性能。	multimodal
2	Enhancing GraphQL Security by Detecting Malicious Queries Using Large Language Models, Sentence Transformers, and Convolutional Neural Networks	提出基于LLM、Sentence Transformer和CNN的GraphQL恶意查询检测方法，提升API安全性。	large language model
3	Modeling Human Responses to Multimodal AI Content	提出MhAIM数据集与T-Lens系统，用于建模人类对多模态AI生成内容的反应，提升LLM的人类感知能力。	multimodal
4	GenOM: Ontology Matching with Description Generation and Large Language Model	GenOM：利用描述生成和大型语言模型进行本体匹配，提升生物医学领域互操作性。	large language model
5	MSRS: Adaptive Multi-Subspace Representation Steering for Attribute Alignment in Large Language Models	提出MSRS，通过多子空间表示引导实现大语言模型中属性对齐，减少属性间的干扰。	large language model
6	Reverse Physician-AI Relationship: Full-process Clinical Diagnosis Driven by a Large Language Model	提出DxDirector-7B，以LLM驱动全流程临床诊断，显著提升诊断准确率并减轻医生负担。	large language model
7	Why Cannot Large Language Models Ever Make True Correct Reasoning?	论文批判：大型语言模型本质限制使其无法实现真正的正确推理	large language model
8	The Knowledge-Reasoning Dissociation: Fundamental Limitations of LLMs in Clinical Natural Language Inference	揭示LLM在临床自然语言推理中知识与推理的解耦局限性	large language model chain-of-thought
9	MCP-Guard: A Multi-Stage Defense-in-Depth Framework for Securing Model Context Protocol in Agentic AI	MCP-Guard：针对Agentic AI中模型上下文协议的多阶段防御框架	large language model
10	Benchmark Dataset Generation and Evaluation for Excel Formula Repair with LLMs	提出Excel公式修复基准数据集生成与评估方法，提升LLM在公式错误纠正中的应用	large language model
11	Who Benefits from AI Explanations? Towards Accessible and Interpretable Systems	针对视觉障碍用户，研究可访问且可解释的AI系统设计方法	multimodal
12	Improving Value-based Process Verifier via Low-Cost Variance Reduction	提出ComMCS方法，通过低成本方差缩减提升基于价值的过程验证器性能	large language model
13	SEQ-GPT: LLM-assisted Spatial Query via Example	提出SEQ-GPT，利用LLM解决空间示例查询（SEQ）中复杂的位置搜索问题。	large language model
14	LeanRAG: Knowledge-Graph-Based Generation with Semantic Aggregation and Hierarchical Retrieval	LeanRAG：基于知识图谱的语义聚合与分层检索生成框架	large language model	✅
15	What to Ask Next? Probing the Imaginative Reasoning of LLMs with TurtleSoup Puzzles	提出TurtleSoup-Bench，用于评估LLM在信息稀疏环境下的想象推理能力。	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	A Curriculum Learning Approach to Reinforcement Learning: Leveraging RAG for Multimodal Question Answering	利用课程学习强化学习，解决检索增强生成多模态问答难题	reinforcement learning curriculum learning large language model
17	We-Math 2.0: A Versatile MathBook System for Incentivizing Visual Mathematical Reasoning	提出We-Math 2.0，通过知识驱动和模型中心的数据建模，提升多模态大语言模型在数学推理上的能力。	reinforcement learning large language model multimodal
18	Is On-Policy Data always the Best Choice for Direct Preference Optimization-based LM Alignment?	揭示DPO中On-Policy数据并非总是最优，提出对齐阶段假设并优化数据选择	DPO direct preference optimization
19	AnalogSeeker: An Open-source Foundation Language Model for Analog Circuit Design	AnalogSeeker：用于模拟电路设计的开源基础语言模型	distillation foundation model	✅
20	Promoting Efficient Reasoning with Verifiable Stepwise Reward	提出基于可验证逐步奖励机制(VSRM)，提升大型推理模型(LRM)的推理效率。	reinforcement learning PPO

🔬 支柱一：机器人控制 (Robot Control) (2 篇)

#	题目	一句话要点	标签	🔗	⭐
21	MCP2OSC: Parametric Control by Natural Language	提出MCP2OSC，利用自然语言提示实现对参数化OSC的精确控制。	OSC large language model
22	Tabularis Formatus: Predictive Formatting for Tables	TaFo：提出一种神经符号方法，用于表格的预测性条件格式设置。	manipulation

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

#	题目	一句话要点	标签	🔗	⭐
23	DINOMotion: advanced robust tissue motion tracking with DINOv2 in 2D-Cine MRI-guided radiotherapy	DINOMotion：基于DINOv2的鲁棒2D-Cine MRI引导放疗组织运动追踪	motion tracking

⬅️ 返回 cs.AI 首页 · 🏠 返回主页