cs.AI（2026-10-22）

📊 共 44 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) (28 🔗2) 支柱二：RL算法与架构 (RL & Architecture) (12) 支柱三：空间感知与语义 (Perception & Semantics) (3 🔗2) 支柱一：机器人控制 (Robot Control) (1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (28 篇)

#题目一句话要点标签🔗⭐
1 Predicting Music Valence and Arousal via Textual and Multimodal In-Context Learning in Large Language Models 提出基于大语言模型的音乐情感预测方法以解决情感价值建模难题 large language model multimodal
2 Measuring the Reasoning-Action Coupling in Driving Vision-Language-Action Models through Counterfactuals 通过反事实方法测量驾驶视觉-语言-行动模型的推理-行动耦合 vision-language-action VLA chain-of-thought
3 Trion-1: Unifying Language, Protein Sequence, and Structure with Diffusion Language Models 提出Trion-1以统一语言、蛋白质序列与结构 foundation model multimodal instruction following
4 Constraint Reasoning and Large Language Models Join Forces: Providing Contrastive Explanations for Real-World Optimisation-Based Applications 结合约束推理与大语言模型提供优化问题的对比解释 large language model
5 Zero-Shot Forecasting of In-Situ Soil Moisture and Temperature with a Time-Series Foundation Model 提出TimesFM模型以实现无数据预测土壤湿度和温度 foundation model
6 Speculative Acceptance Does Not Decay with Position: A Full-Length Diagnostic on Long Chain-of-Thought with Independent Drafts 提出全长诊断方法以解决链式思维接受率衰减问题 chain-of-thought
7 Large language models show systematic failures in incorporating evolving medical knowledge 提出ChronoEvo以解决大型语言模型在医学知识更新中的不足 large language model
8 PointerBench: A Diagnostic Benchmark for Fine-Grained GUI Grounding 提出PointerBench以解决细粒度GUI定位问题 large language model multimodal ✅
9 AgenticEvolve: Evolutionary Program Search with Persistent Language Model Agents that Design and Run Experiments 提出AgenticEvolve以提升进化程序搜索效率 large language model
10 VERA: A Responsible AI Dashboard for Assessing LLM Deployments in Light of the EU AI Act 提出VERA以促进大型语言模型的负责任AI评估 large language model
11 Who Is Sycophantic to Whom? Confident LLMs, Human Deference, and the Collapse of Trust Calibration 提出人类依赖指数以解决信任校准崩溃问题 large language model
12 Social-HUMINT: An LLM-Augmented Unified Framework for Measuring Social Media Exposure 提出Social-HUMINT框架以量化社交媒体曝光风险 large language model
13 Contestable Agents for Data Governance: Recording Human Stewardship of AI-Generated Metadata Across Structured and Unstructured Systems 提出可争议代理以解决数据治理中的元数据管理问题 large language model
14 Certified Execution Harness for Agentic AI: Safe and Least-Intrusive Petri-Net Supervisory Control Between Reasoning and Action 提出安全工具执行框架以解决大语言模型代理的安全性问题 large language model
15 ValuSpec: Plug-and-Play Candidate Valuation before Target Verification for Tree-Based Speculative Decoding 提出ValuSpec以解决树形结构候选分支验证效率低下问题 large language model
16 Network topology shapes consensus and vulnerability in LLM populations 探讨网络拓扑对大语言模型共识与脆弱性的影响 large language model
17 Can AI Co-Design Structures? An AI Agent for Early-Stage Structural Design 提出AI代理以实现早期结构设计的协同设计 large language model
18 Directing to Diversify: Multi-Directional Search for Automatic Heuristic Design 提出多方向搜索框架以自动化启发式设计 large language model
19 Evaluating Artificial Wisdom (AW): A Trajectory-Based Framework for Context-Sensitive Decision Regulation 提出基于轨迹的框架以评估人工智慧的决策调节能力 large language model
20 SecLog-Bench: A Scalable Benchmark for Cybersecurity Log Reasoning with Scaffolded Evidence 提出SecLog-Bench以解决网络安全日志推理的规模化问题 large language model
21 Low-Latency MoE Inference in Space Data Centers: A Wireless-Power-Transfer Perspective 提出SpaceWPT框架以解决空间数据中心的低延迟MoE推理问题 large language model
22 Harnessing Language Models as Optimization Modeling Agents 提出OPTICYCLE和OPTISTATE以优化语言模型的建模能力 large language model
23 ProQuant: Preserving Expert Quality over Quantity in Ultra-Low-Bit Quantization of Mixture-of-Experts Models 提出ProQuant以解决MoE模型超低比特量化中的质量保持问题 large language model ✅
24 Abstract Priorities, Concrete Rewards: How Stated Preferences Relate to What People Reward in AI Responses 探讨人类偏好与AI响应奖励之间的关系 large language model
25 AgentDP: Towards Automated Model Deployment for Heterogeneous Embedded AI Applications 提出AgentDP以解决异构嵌入式AI应用的自动化模型部署问题 large language model
26 Faithful Latent Harm Monitors Work No Better Than the Models Own Behaviour at a Matched False Alarm Rate 提出改进的潜在安全监测方法以提升有害请求检测能力 large language model
27 TACS-RF: Tool-driven Agentic Circuit Synthesis of RF and mmWave Modules 提出TACS-RF框架以实现RF和毫米波模块的自动化设计 large language model
28 Mining the Design Space of Multi-Agent Workflows 提出一种新方法以挖掘多智能体工作流设计原则 large language model

🔬 支柱二：RL算法与架构 (RL & Architecture) (12 篇)

#题目一句话要点标签🔗⭐
29 An Introduction to World Models for Generalist Agents 提出统一的世界模型框架以支持通用智能体的决策能力 reinforcement learning policy learning world model
30 Dual Exam: Controlled Evaluation of Multimodal Agents and World Models 提出Dual Exam以评估多模态智能体与世界模型的交互 world model world models multimodal
31 Beyond Phenotypic Similarity: Relation Refinement and Selective Population Graph Learning for Multimodal Brain Disorder Diagnosis 提出PopDRP框架以优化多模态脑部疾病诊断 representation learning multimodal
32 Anticipating Social Consequences with Distributional World Models 提出分布式社会世界模型以解决社交互动预测问题 world model world models
33 RIPPLE: Recovery Internalization through Process-Level Policy Learning and Effect-Aware Credit for Long-Horizon Agentic Coding 提出RIPPLE框架以解决长时间编码代理的恢复问题 reinforcement learning policy learning distillation
34 DIAL: Steerable Language Models with Distilled Instruction-Adherence Levels 提出DIAL以实现可控的语言模型指令遵循 distillation classifier-free guidance instruction following
35 GateFormer: Forecast-Space Decomposition for Intra-Hour Solar Irradiance Prediction from Sky Images 提出GateFormer以解决基于天空图像的太阳辐射预测问题 MAE multimodal
36 Conflict-Aware Logit Adapters for Utility-Preserving Anti-Distillation 提出冲突感知逻辑适配器以解决蒸馏训练中的信息泄露问题 distillation
37 PhysTrace: Interpretable Temporal Evaluation of Physical Consistency in Video Generative Models 提出PhysTrace以解决视频生成模型物理一致性评估问题 world action model world action models
38 Reward-Weighted Preference Optimization for Selective Combinatorial Optimization Problems 提出奖励加权偏好优化以解决选择性组合优化问题 reinforcement learning preference learning
39 Agentic Knowledge-Model Co-Evolution via Consistency Regularization 提出SPLICE框架以解决知识模型共演化问题 reinforcement learning distillation
40 Beyond On-Policy vs. Off-Policy: At What Level Should LLM Agent Distillation Follow the Student? 提出层次化蒸馏框架以优化LLM代理训练 distillation

🔬 支柱三：空间感知与语义 (Perception & Semantics) (3 篇)

#题目一句话要点标签🔗⭐
41 Adaptive Skill Granularity for Coding Agents: Runtime Routing and Macro-to-Micro Recovery 提出自适应技能粒度框架以解决编码代理的执行控制问题 affordance large language model
42 Mitigating Saliency Collapse: Robust Saliency-Aware Long-Text Image-Text Alignment 提出RoSA框架以解决长文本图像-文本对齐中的显著性崩溃问题 scene understanding ✅
43 Mitigating Saliency Collapse: Robust Saliency-Aware Long-Text Image-Text Alignment 提出RoSA框架以解决长文本图像-文本对齐中的显著性崩溃问题 scene understanding ✅

🔬 支柱一：机器人控制 (Robot Control) (1 篇)

#题目一句话要点标签🔗⭐
44 Knowing Whom to Trust is not Enough: The Trust--Agreement Gap in Repeated Multi-Party LLM Negotiation 提出信任评估模型以解决多方LLM谈判中的信任与协议差距问题 manipulation large language model

⬅️ 返回 cs.AI 首页 · 🏠 返回主页