cs.AI（2026-07-30）

📊 共 49 篇论文 | 🔗 5 篇有代码

🎯 兴趣领域导航

支柱九：具身大模型 (Embodied Foundation Models) (29 🔗1) 支柱二：RL算法与架构 (RL & Architecture) (15 🔗4) 支柱一：机器人控制 (Robot Control) (2) 支柱八：物理动画 (Physics-based Animation) (1) 支柱七：动作重定向 (Motion Retargeting) (1) 支柱三：空间感知与语义 (Perception & Semantics) (1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (29 篇)

#题目一句话要点标签🔗⭐
1 PathView-Bench: Can Multimodal Large Language Models Achieve Fine-grained Multiscale Understanding of Pathology Images? 提出PathVU基准以解决病理图像多尺度理解问题 large language model multimodal
2 IndustryForge-27B: A Domain-Enhanced Multimodal Foundation Model for Industrial CAD 提出IndustryForge-27B以解决工业CAD设计自动化问题 foundation model multimodal
3 IFHierBench: Hierarchical Instruction Following for Large Language Models 提出IFHierBench以解决层次化指令遵循问题 large language model instruction following
4 AISPA: User-Centric System Prompt Auditing for Large Language Model Applications 提出AISPA框架以审计大型语言模型应用中的系统提示 large language model foundation model
5 DualG-MRAG: Decoupling Macro-Reasoning and Micro-Matching for Multimodal Retrieval-Augmented Generation 提出DualG-MRAG以解决多模态检索增强生成中的复杂推理问题 multimodal
6 A report-grounded vision-language foundation model for colonoscopy from 280000 routine reports 提出EndoCLIP以解决内窥镜报告与图像关联不足问题 foundation model
7 Asymmetric Communication: Large Language Models and Language Games 提出不对称沟通框架以重新审视语言模型的属性 large language model
8 Distilling Answer Set Programming Theories from Large Language Models 提出神经符号方法从大型语言模型中蒸馏答案集编程理论 large language model
9 MMLDSum-LLM: Multimodal Long-Document Summarization with Visual-Alignment and Keyword-Aware 提出MMLDSum-LLM以解决多模态长文档摘要中的信息遗漏问题 multimodal
10 Specification-Guided Synthesis of Deadlock-Free Communication Protocol Refinements with Large Language Models 提出Syntropy框架以解决通信协议死锁问题 large language model
11 One Anchor for All: Unified Multilingual and Multimodal Safety Alignment for LVLMs 提出统一多语言多模态安全对齐框架以应对复合攻击问题 multimodal
12 Guiding Large Language Models with Genetic Programming-Evolved Heuristic Knowledge for Dynamic Multi-Mode Project Scheduling 利用遗传编程进化的启发式知识指导大型语言模型进行动态多模式项目调度 large language model
13 Using Large Language Models for Idea Generation in Innovation 利用大型语言模型提升创新产品构思的有效性 large language model
14 MMHBench: A Multi-Perspective Benchmark for Mental Health Understanding in Long-Form Videos 提出MMHBench以解决长视频心理健康理解问题 large language model multimodal
15 Old Tricks, New Models: How Simple Image Transformations Break Modern AI-based Content Moderation 提出简单图像变换以突破现代AI内容审核系统的局限 foundation model multimodal
16 Albilich: Steerable Proof-State Orchestration for LLM-Based Mathematical Research with CAS Integration 提出Albilich以解决数学研究中的长远证明协调问题 large language model
17 Strategy, Not Payoffs: A Behavioural Embedding of Normal-Form Games 提出行为嵌入以解决LLM战略能力转移问题 large language model
18 MANTA: Multi-Agent Network Topology Adaptation for Self-Evolving Multi-Agent Systems 提出MANTA框架以解决多智能体系统通信拓扑固定性问题 large language model
19 A Fuzzy Rule-based Neuro-Symbolic Approach for Pipe Severity Prediction in Sewer Networks 提出模糊规则神经符号方法以解决污水管道严重性预测问题 large language model
20 HyperClaim: Fine-Grained Cross-Modal Hypergraph Reasoning for Video Misinformation Detection 提出HyperClaim以解决视频虚假信息检测中的多模态推理问题 multimodal
21 From Textual Requirements to Microservice Architectures - A Comprehensive Evaluation of LLM-Based Design Synthesis 基于LLM的设计合成方法解决微服务架构识别问题 large language model
22 MemHarness: Memory Is Reconstructed, Not Replayed 提出MemHarness框架以重构而非重放记忆提升LLM智能体表现 large language model
23 LLM-Guided Evolutionary Search for Constraint Model Reformulation to Improve Solver Efficiency 提出基于LLM的进化搜索以优化约束模型重构 large language model
24 Vibe-FDTR: An agent-oriented framework for reproducible frequency-domain thermoreflectance data analysis 提出Vibe-FDTR框架以解决频域热反射数据分析的复杂性问题 large language model
25 An Instrument to Evaluate Governance Proposals: AI Policy Analysis at Scale 提出AI政策分析框架以评估治理提案 large language model
26 DataClawEval: A Benchmark for Data Engineering Agents in Real Industrial Harness 提出DataClawEval基准以评估工业数据工程代理的能力 large language model ✅
27 Safeguards Based on Copyable Context Cannot Provide Reliable Safety for LLMs 提出可信凭证以解决大型语言模型安全性问题 large language model
28 ARES: Adaptive Reasoning-Effort Steering for PPA- and Cost-Aware RTL Optimization with LLM Agents 提出ARES以优化RTL设计的PPA和成本问题 large language model
29 MemTxn: A Transaction Boundary for Source-Supported Updates and Complete-State Recovery in Agent Memory 提出MemTxn以解决持久内存更新和状态恢复问题 large language model

🔬 支柱二：RL算法与架构 (RL & Architecture) (15 篇)

#题目一句话要点标签🔗⭐
30 MARS-RA: Rank Aggregation for Credit Assignment via Multimodal Comparisons in Embodied Multi-Agent Cooperation 提出MARS-RA框架以解决多智能体合作中的信用分配问题 reinforcement learning reward shaping embodied AI
31 World Action Planner: Generalizable Decision-Making with Action-Conditioned World Models 提出世界行动规划器以解决通用决策问题 imitation learning world model world models
32 Correcting What You Cannot See: Credit Assignment for Perception Distillation in Multimodal Reasoners 提出感知纠正蒸馏方法以解决多模态推理中的奖励分配问题 distillation multimodal
33 Group-Reflective Self-Distillation for Agentic Reinforcement Learning 提出群体反思自蒸馏以解决强化学习中的稀疏监督问题 reinforcement learning distillation large language model
34 Tycho: Active Abstraction with Programmatic World Models for ARC-AGI-3 提出Tycho以解决ARC-AGI-3中的主动抽象问题 world model world models
35 SKILL-KD: Contrastive Skill Distillation for LLM Agents 提出SKILL-KD以解决LLM代理技能获取不足的问题 distillation large language model
36 From Scoring to Acting: Outcome-Verified Comparative Self-Distillation for LLM Agents 提出结果验证的比较自蒸馏方法以提升LLM代理能力 teacher-student distillation
37 Not All Tokens Deserve Equal Credit: Counterfactual Sensitivity Credit Reallocation for Long-CoT Reasoning 提出反事实敏感性信用重新分配以提升长链推理效果 reinforcement learning distillation large language model
38 ReDiPPO: Reference-Guided Value Calibration and Discrepancy-Aware Token Reweighting for Mathematical Reasoning 提出ReDiPPO框架以解决数学推理中的价值评估问题 reinforcement learning PPO large language model ✅
39 SVR: Self-Verifying Refinement via Joint Verdict-Confidence Reinforcement Learning for Adaptive Test-Time Compute 提出自验证精炼方法以优化自适应测试时计算 reinforcement learning
40 When Derived Measurements Mislead: Quantifying and Mitigating LLM Over-Trust with Privileged-Modality Reliability Evidence 提出DFOT框架以解决LLM对派生测量的过度信任问题 distillation large language model ✅
41 Interpretable Representation via LLM-Driven Generative Disentanglement for Local-Life Service Recommendation 提出LGRID以解决SID生成中的语义纠缠问题 representation learning large language model
42 The Geometric Nature and a Free Proxy for Flow-Matching Uncertainty 提出几何不确定性代理以解决流匹配模型的信任问题 flow matching ✅
43 Hierarchical Latent Reasoning for LLM-based Recommendation 提出HiLaR框架以优化LLM推荐系统中的潜在推理 representation learning large language model ✅
44 DeepResearch Agent System 提出DeepResearch Agent System以解决深度信息检索与自主研究问题 reinforcement learning large language model

🔬 支柱一：机器人控制 (Robot Control) (2 篇)

#题目一句话要点标签🔗⭐
45 Security of World-Model-Based Embodied AI: A Lifecycle of Threats, Defenses, and Evaluation 提出全面生命周期安全框架以应对世界模型基础的具身AI威胁 manipulation world model world models
46 QuantWAMs: Calibrating at the Right Granularity for World Action Models 提出QuantWAMs框架以优化世界动作模型的量化过程 manipulation world action model world action models

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

#题目一句话要点标签🔗⭐
47 A foundation model of numerical intelligence with cross-disciplinary generalization 提出UNICON以解决跨学科数值智能问题 UniCon large language model foundation model

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

#题目一句话要点标签🔗⭐
48 SpatialCLI: Learning to Reason With Spatial Tools, Then Without Them 提出SpatialCLI框架以解决视觉语言模型的空间推理问题 spatial relationship

🔬 支柱三：空间感知与语义 (Perception & Semantics) (1 篇)

#题目一句话要点标签🔗⭐
49 Integrating AI into Requirements Quality Learning in Software Engineering Education: A TPACK-Guided Empirical Study 提出TPACK指导的多代理AI工具以提升软件工程教育中的需求质量学习 affordance

⬅️ 返回 cs.AI 首页 · 🏠 返回主页