cs.AI（2026-07-05）

📊 共 78 篇论文

🎯 兴趣领域导航

支柱九：具身大模型 (Embodied Foundation Models) (49) 支柱二：RL算法与架构 (RL & Architecture) (24) 支柱四：生成式动作 (Generative Motion) (3) 支柱一：机器人控制 (Robot Control) (1) 支柱八：物理动画 (Physics-based Animation) (1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (49 篇)

#题目一句话要点标签🔗⭐
1 Multimodal Knowledge Edit-Scoped Generalization for Online Recursive MLLM Editing 提出ScopeEdit以解决多模态知识编辑的语义边界问题 large language model multimodal
2 Revisiting Chain-of-Thought Reasoning under Limited Supervision: Semi-supervised Chain-of-Thought Learning 提出Semi-CoT以解决有限监督下的推理能力激活问题 large language model chain-of-thought
3 MMIR-TCM: Memory-Integrated Multimodal Inference and Retrieval for TCM Clinical Decision Support 提出MMIR-TCM以解决中医诊断中的主观性与可重复性问题 large language model multimodal
4 InduceKV: Fixed-Footprint Continual Adaptation of Multimodal LLMs via Inducing KV Memories 提出InduceKV以解决多模态LLMs的持续适应问题 large language model multimodal
5 Hidden Forgetting in Continual Multimodal Learning: When Accuracy Survives but Grounding Fails 提出RCL框架以解决多模态学习中的隐性遗忘问题 large language model multimodal
6 Pre-Flight: A Benchmark for Evaluating Large Language Models on Aviation Operational Knowledge 提出Pre-Flight基准以评估航空运营知识中的大型语言模型 large language model
7 OntoLearner: A Modular Python Library for Ontology Learning with Large Language Models 提出OntoLearner以解决本体学习的系统性评估问题 large language model
8 A Practice Auditing Framework for Large Language Model Use: Collective Empiricism, Pseudo-Rational Cognition, and Governance of AI-Generated Content 提出实践审计框架以治理大型语言模型生成内容 large language model
9 Psychological Imagination Networks Show Cross-Population Centrality and Clustering Alignment in Humans That Large Language Models Fail to Replicate 提出心理想象网络以解决人类与大型语言模型的想象结构差异问题 large language model
10 Leveraging Metamemory Agent for Enhanced Data-Free Code Generation in Large Language Models 提出变记忆体代理以解决无数据代码生成问题 large language model
11 Exploring Large Language Models for Access Control Policy Synthesis and Summarization 利用大型语言模型自动生成和分析访问控制策略 large language model
12 Search for Truth from Reasoning: A Dynamic Representation Editing Framework for Steering LLM Trajectories 提出DynaSteer框架以优化大语言模型推理过程 large language model chain-of-thought
13 Spatial Reasoning via Modality Switching Between Language and Symbolic Representation 提出多模态切换方法以提升空间推理能力 large language model multimodal
14 BaRA: Budget-constrained and Reliable Web Data Collection Agent 提出BaRA以解决预算限制下的可靠网页数据收集问题 large language model multimodal
15 From Actions to Understanding: Conformal Interpretability of Temporal Concepts in LLM Agents 提出顺序一致性框架以提升LLM代理的可解释性 large language model
16 Generative AI and Federated Learning for Intrusion Detection Systems: A Survey 综述生成式AI与联邦学习在入侵检测系统中的应用与挑战 large language model
17 CreativityNeuro: Steering Language Model Weights to Improve Divergent Thinking and Reduce Mode Collapse 提出CreativityNeuro以提升语言模型的发散思维能力 large language model
18 AgenticDataBench: A Comprehensive Benchmark for Data Agents 提出AgenticDataBench以解决数据代理评估不足的问题 large language model
19 Fast Multi-dimensional Refusal Subspaces via RFM-AGOP 提出RFM-AGOP以快速识别多维拒绝子空间 large language model
20 Gravity-Awareness: Deep Learning Models and LLM Simulation of Human Awareness in Altered Gravity 提出重力感知模型以解决人类在改变重力环境中的适应问题 large language model
21 EO-Agents: A Three-Agent LLM Pipeline for Earth Observation Hypothesis Generation 提出EO-Agents以解决地球观测假设生成问题 large language model
22 Path-level Hindsight Instructions for Semantic Exploration in Vision-Language Navigation 提出Phi-Nav以解决视觉语言导航中的语义探索问题 VLN
23 SimWorlds: A Multi-Agent System for Dynamic 3D Scene Creation 提出SimWorlds以解决动态3D场景生成问题 embodied AI
24 Safety Targeted Embedding Exploit via Refinement 提出STEER以解决多语言安全机制泛化不足问题 large language model
25 ContextSniper: AntTrail's Token-Efficient Code Memory for Repository-Level Program Repair 提出ContextSniper以解决代码修复中的上下文效率问题 large language model
26 UA-ChatDev: Uncertainty-Aware Multi-Agent Collaboration for Reliable Software Development 提出UA-ChatDev以解决软件开发中的不确定性问题 large language model
27 ReContext: Recursive Evidence Replay as LLM Harness for Long-Context Reasoning 提出RECONTEXT以解决长上下文推理中的证据利用问题 large language model
28 AI Assistance for Human Review of Default Judgments 提出默认判决审查的AI助手以提高法院效率 large language model
29 ADVENT: LLM-Driven Automatic Predicate Invention for ILP 提出ADVENT以解决归纳逻辑编程中的谓词发明问题 large language model
30 Robust for the Wrong Reasons: The Representational Geometry of LLM Robustness to Science Skepticism 探讨大型语言模型对科学怀疑的鲁棒性及其表现机制 large language model
31 Prompt Coverage Adequacy 提出Prompt Coverage Adequacy以解决LLM驱动软件开发测试问题 large language model
32 Behind the Refusal: Determining Guardrail Activation via Behavioral Monitoring 提出黑箱守卫侦察方法以解决AI系统安全监测问题 large language model
33 SelectTSL: Prompt-Guided Selective Target Sound Localization in Complex Scenarios 提出SelectTSL以解决复杂场景下的目标声音定位问题 multimodal
34 SkillFuzz: Fuzzing Skill Composition for Implicit Intents Discovery in Open Skill Marketplaces 提出SkillFuzz以解决隐式意图发现问题 large language model
35 TestEvo-Bench: An Executable and Live Benchmark for Test and Code Co-Evolution 提出TestEvo-Bench以解决测试与代码共同演化问题 foundation model
36 ADMC: Attention-based Diffusion Model for Missing Modalities Feature Completion 提出ADMC以解决多模态缺失特征补全问题 multimodal
37 Formal Semantics for Agentic Tool Protocols: A Process Calculus Approach 提出过程演算形式化以验证代理工具协议的有效性 large language model
38 The Token Not Taken: Sampling, State, and the Stochasticity of AI Agents 提出分层分析方法以解决AI代理系统的随机性问题 foundation model
39 Beyond the Library: An Agentic Framework for Autoformalizing Research Mathematics 提出代理框架以实现研究数学的自动形式化 large language model
40 ClawArena-Team: Benchmarking Subagent Orchestration and Dynamic Workflows in Language-Model Agents 提出ClawArena-Team基准以评估语言模型代理的管理能力 multimodal
41 RedCoder: Automated Multi-Turn Red Teaming for Code LLMs 提出RedCoder以解决代码生成模型的安全性评估问题 large language model
42 Who Gets the Reward & Who Gets the Blame? Evaluation-Aligned Training Signals for Multi-LLM Agents 提出统一的训练信号以优化多智能体系统中的LLM表现 large language model
43 It's a TRAP! Task-Redirecting Agent Persuasion Benchmark for Web Agents 提出TRAP基准以研究网络代理的任务重定向问题 large language model
44 From Experiments to Expertise: Scientific Knowledge Consolidation for AI-Driven Computational Physics 提出QMatSuite以解决AI驱动计算物理中的知识积累问题 large language model
45 ChemGraph-XANES: An Agentic Framework for XANES Simulation and Curation 提出ChemGraph-XANES框架以简化XANES模拟与分析 large language model
46 BLAgent: Agentic RAG for File-Level Bug Localization 提出BLAgent以解决文件级别的Bug定位问题 large language model
47 KernelSight-LM: A Kernel-Level LLM Inference Simulator 提出KernelSight-LM以解决LLM推理性能评估问题 large language model
48 Four Types of LLM Reliance and Their Predictors Among Undergraduate Writers: A Mixed-Methods Study at a Minority-Serving R1 University 提出四种LLM依赖类型以改善本科生写作支持 large language model
49 Translating Natural Language to Strategic Temporal Specifications via LLMs 提出基于大语言模型的自然语言到战略时序规范的翻译框架 large language model

🔬 支柱二：RL算法与架构 (RL & Architecture) (24 篇)

#题目一句话要点标签🔗⭐
50 Safe and Adaptive Cloud Healing: Verifying LLM-Generated Recovery Plans with a Neural-Symbolic World Model 提出PASE框架以解决云服务自愈问题 reinforcement learning deep reinforcement learning DRL
51 Physically Native World Models: A Hamiltonian Perspective on Generative World Modeling 提出哈密顿世界模型以解决物理可靠性不足的问题 reinforcement learning world model world models
52 Enhancing Fitness Intelligence through Domain-Specific LLM Post-Training 提出FitOne以解决科学健身教练领域的知识不足问题 reinforcement learning large language model foundation model
53 OPINE-World: Programmatic World Modeling with Ontology-error-Prioritized Interactive Exploration 提出OPINE-World以解决程序合成世界建模的挑战 world model world models
54 Scaling with Confidence: Calibrating Confidence of LLMs for Adaptive Test Time Scaling 提出C3RL与CAS以解决LLMs信心校准问题 reinforcement learning reward design large language model
55 DRL-CLBA: A Clean Label Backdoor Attack for Speech Classification via DDPG Reinforcement Learning 提出DRL-CLBA以解决语音分类中的清标签后门攻击问题 reinforcement learning DRL
56 Repair the Amplifier, Not the Symptom: Stable World-Model Correction for Agent Rollouts 提出WM-SAR以解决大规模规划图中的错误修复问题 world model world models
57 Grounded Iterative Language Planning: How Parameterized World Models Reduce Hallucination Propagation in LLM Agents 提出基于参数化世界模型的迭代语言规划以减少LLM代理的幻觉传播 world model world models
58 Ophiuchus: Incentivizing Tool-augmented "Think with Images" for Joint Medical Segmentation, Understanding and Reasoning 提出Ophiuchus以解决医疗图像细粒度分析问题 reinforcement learning multimodal chain-of-thought
59 Purified OPSD: On-Policy Self-Distillation Without Losing How to Think 提出Purified OPSD以解决长链推理模型的自蒸馏问题 distillation chain-of-thought
60 GR2 Technical Report 提出GR2框架以优化工业推荐系统的重排序问题 reinforcement learning distillation reward design
61 Procedural Memory Distillation: Online Reflection for Self-Improving Language Models 提出程序记忆蒸馏以提升自我改进语言模型的性能 reinforcement learning distillation
62 Hardware-Enforced Semantic Coordination for Safety-Critical Real-Time Autonomous Systems 提出硬件强制语义协调架构以解决安全关键实时系统的协调问题 world model world models large language model
63 MAGIK: Mapping to Analogous Goals via Imagination-enabled Knowledge Transfer 提出MAGIK框架以实现类比任务的知识迁移 reinforcement learning zero-shot transfer
64 Beyond Next-Token Prediction: An RLVR Proof of Concept for Tool-Use Agents on Atlassian Workflows 提出RLVR方法以优化企业API工具使用的效果 reinforcement learning large language model
65 Scaling Trends for Lie Detector Oversight in Preference Learning 提出可扩展的谎言检测器监督方法以解决偏好学习中的欺骗行为监控问题 preference learning
66 Hawk: Harnessing Hardware-Aware Knowledge for High-Performance NPU Kernel Generation 提出Hawk框架以解决NPU内核生成中的硬件约束问题 distillation large language model
67 A Hippocampus for Linear Attention: An Exact Memory for What the Recurrent State Forgets 提出HOLA模型以解决线性注意力的记忆丢失问题 linear attention
68 DRIFTLENS: Measuring Memory-Induced Reasoning Drift in Personalized Language Models 提出DRIFTLENS以测量个性化语言模型中的推理漂移 DPO instruction following
69 ExPerT: Personalizing LLM Responses to Users' Domain Expertise via Query-Wise Semantic and Keystroke Behavioral Cues 提出ExPerT框架以解决LLM个性化响应问题 MAE large language model
70 LLMs as Teaching Assistants for Mathematics Exam Grading: Reliability, and Practical Usability 评估大型语言模型作为数学考试评分助手的可靠性与实用性 MAE large language model
71 Playing 20 Question Game with Policy-Based Reinforcement Learning 提出基于策略的强化学习方法以优化20问游戏中的提问策略 reinforcement learning
72 Hybrid Diffusion Transformer for Instruction-Guided Audio Editing via Rectified Flow 提出混合扩散变换器以解决指令引导音频编辑问题 flow matching multimodal
73 Active Sensing for RIS-Aided Tracking and Power Control: A Hybrid Neuroevolution and Supervised Learning Approach 提出一种混合神经进化与监督学习的方法以解决RIS辅助的跟踪与功率控制问题 reinforcement learning deep reinforcement learning

🔬 支柱四：生成式动作 (Generative Motion) (3 篇)

#题目一句话要点标签🔗⭐
74 Text-Driven 3D Indoor Scene Synthesis in Non-Manhattan Environments 提出SPG-Layout以解决非曼哈顿环境中的3D室内场景合成问题 physically plausible spatial relationship large language model
75 An Efficient vLLM-Based Inference Pipeline for Unified Audio Understanding and Generation 提出基于vLLM的推理管道以解决多模态生成问题 classifier-free guidance multimodal
76 Low-Latency Task-Oriented Image Transmission with Opportunistic Spectrum Access 提出低延迟任务导向图像传输框架以解决频谱有限问题 VQ-VAE

🔬 支柱一：机器人控制 (Robot Control) (1 篇)

#题目一句话要点标签🔗⭐
77 Automated grading of Linux/bash examinations using large language models: a four-level cognitive taxonomy approach 利用大型语言模型实现Linux/bash考试的自动评分 manipulation MAE large language model

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

#题目一句话要点标签🔗⭐
78 Scene-Conditioned PINN-GNN for Multipath RF Maps: Cross-Scene Generation and In-Scene Completion 提出基于PINN-GNN的RF地图构建框架以解决多路径传播问题 PULSE

⬅️ 返回 cs.AI 首页 · 🏠 返回主页