cs.AI（2026-08-17）

📊 共 31 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) (17 🔗3) 支柱二：RL算法与架构 (RL & Architecture) (8 🔗1) 支柱一：机器人控制 (Robot Control) (5) 支柱三：空间感知与语义 (Perception & Semantics) (1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (17 篇)

#	题目	一句话要点	标签	🔗	⭐
1	CACSurv: Concordance-Aligned Comparative Learning with Large Language Models for Cancer Survival Prediction	提出CACSurv以解决癌症生存预测中的不匹配问题	large language model multimodal	✅
2	HaReCAP: Habitual-action Grounding for Recursive Large Language Model Agents	提出HaReCAP以解决长时间任务中的最后一步动作基础问题	large language model
3	Measuring Obedience to Authority Across Large Language Models with the Milgram Paradigm	将米尔格拉姆服从实验引入大型语言模型以评估其服从性	large language model
4	MUPA$^{2}$E: Multimodal Unified Perception with Asymmetric Attention for Emotion Assessment	提出MUPA²E框架以解决多模态情感评估问题	multimodal
5	ALPS: Measuring Valid Creativity in Large Language Models with Mathematical Construction	提出ALPS基准以衡量大型语言模型的有效创造力	large language model
6	LAVA: Logic-Aware Validation and Augmentation Framework for Large-Scale Financial Document Auditing	提出LAVA框架以解决金融文档审计中的验证与增强问题	large language model multimodal
7	Topological Attribution Distance (TAD): Revealing Segment-Level RAG Influence on LLM Output Geometry for Incident Log Analysis	提出拓扑归因距离（TAD）以解决LLM输出可信度问题	large language model
8	TDD-Agent: Test-Driven Reasoning for Code Generation	提出TDD-Agent以解决代码生成中的正确性问题	large language model
9	Reconstruction: A Blind Benchmark for Recovering Research Ideas from Pre-Publication Bibliographies	提出重建基准以从预出版文献中恢复研究思想	large language model
10	Listen, Reason, and Segment: Aligning LALMs with Editorial Judgment for Media Chapterization	提出AudioChaps框架以解决音频章节化问题	chain-of-thought	✅
11	A Policy Algebra for Trust-Preserving Agentic AI Execution	提出信任保护的代理AI执行策略代数以解决可靠性问题	large language model
12	Think Inside the Chunk: RegulaRAG for Regulation-Compliant Scenario Generation using LLMs: A Case Study of UN Regulation No. 152	提出RegulaRAG以解决合规场景生成问题	large language model
13	Ventor-QTest: Threat-Model-Driven Verification of Vendor-Hosted LLM APIs	提出Ventor-QTest以解决第三方LLM API审核问题	large language model	✅
14	AeroCopilotBench: A Two-Tier Benchmark for Evaluating LLM Agents as Aviation Copilots in an Interactive Virtual Cockpit Environment	提出AeroCopilotBench以解决航空领域LLM代理评估问题	large language model
15	Revisiting the Performance of Generative Artificial Intelligence on Introductory Object-Oriented Programming Assessments: Insights from 2026	评估生成式人工智能在面向对象编程评估中的表现	large language model
16	Decoupled Temporal Encoding for Generative Recommendation	提出解耦时间编码以解决推荐系统中的时间动态问题	TAMP
17	MUSE: An Interactive Meta-Agent for Understanding and Steering LLM-powered Data Science Systems	提出MUSE以增强用户对数据科学系统的理解与控制	large language model

🔬 支柱二：RL算法与架构 (RL & Architecture) (8 篇)

#	题目	一句话要点	标签	🔗	⭐
18	FabriMAE I Trust Myself? Self-Evaluating VLA Action Generation with Markov Attention Entropy	提出MAE框架以解决VLA自我评估可靠性问题	MAE vision-language-action VLA
19	Hypergraph-based Multimodal Retrieval-Augmented Generation with Incremental Refinement	提出Hyper-M2RAG以解决多模态文档检索中的高阶关联问题	representation learning multimodal	✅
20	Large language models as synthetic clinical experts to inform longitudinal rare-disease modeling	利用大型语言模型提升罕见疾病的纵向建模精度	representation learning large language model
21	DriveCache: Action-Aware Caching for Driving World Model Inference	提出DriveCache以解决驾驶世界模型推理中的缓存效率问题	world model world models
22	Offline Reinforcement Learning for Hemodynamic Management of Sepsis in the ICU: a MIMIC-IV Study with Dual Off-Policy Evaluation	提出离线强化学习以优化ICU中脓毒症的血流动力学管理	reinforcement learning offline reinforcement learning
23	A Scalable Pipeline for LLM-Teacher Distillation Labeling: Work-Stealing Job Scheduling and Memory-Aware GPU Concurrency	提出可扩展的LLM教师蒸馏标注管道以解决GPU工作负载调度问题	distillation
24	Chronocooked: A Benchmark for Implicit Interval Timing in Reinforcement Learning Agents	提出Chronocooked基准以研究强化学习中的隐式时间间隔问题	reinforcement learning
25	A Shop Floor Production Scheduling Case based on RFID-supported Smart Factory	基于RFID技术提出动态车间生产调度方法以应对不确定性问题	reinforcement learning deep reinforcement learning

🔬 支柱一：机器人控制 (Robot Control) (5 篇)

#	题目	一句话要点	标签	🔗	⭐
26	Trajectory-Level Automatic Curriculum Learning for Legged Locomotion on Unstructured Terrain	提出轨迹级自动课程学习以解决复杂地形行走问题	legged locomotion locomotion locomotion policy
27	DeepInsight II: One Trace from Benchmark to Robot	提出DeepInsight II以解决机器人评估与部署风险问题	whole-body control manipulation sim-to-real
28	Beyond Asking: A Pipeline for Personalized Game Generation that Reads Players from Behavior	提出个性化游戏生成管道以解决玩家行为理解问题	manipulation large language model
29	Digital Twin Degradation: Detecting Cyber Physical Attacks via Temporal Inconsistencies	提出基于时间一致性检测的数字双胞胎降级攻击识别方法	manipulation
30	Prior Audit-Repair Context Shifts LLM Verifier Thresholds Toward Leniency	提出审计-修复上下文以降低LLM验证器的阈值	manipulation

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

#	题目	一句话要点	标签	🔗	⭐
31	Cross-Sign Language Transfer Learning Using Domain Adaptation with Multi-scale Temporal Alignment	提出跨手语迁移学习方法以解决手语识别资源不足问题	optical flow

⬅️ 返回 cs.AI 首页 · 🏠 返回主页