cs.AI（2026-03-23）

📊 共 26 篇论文 | 🔗 1 篇有代码

🎯 兴趣领域导航

支柱九：具身大模型 (Embodied Foundation Models) (15) 支柱二：RL算法与架构 (RL & Architecture) (9 🔗1) 支柱六：视频提取与匹配 (Video Extraction) (1) 支柱五：交互与反应 (Interaction & Reaction) (1)

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

#	题目	一句话要点	标签	🔗	⭐
1	Structured Visual Narratives Undermine Safety Alignment in Multimodal Large Language Models	ComicJailbreak：利用结构化视觉叙事攻击多模态大语言模型的安全对齐	large language model multimodal
2	A Multidisciplinary AI Board for Multimodal Dementia Characterization and Risk Assessment	Cerebra：多模态AI协作系统，用于痴呆症特征分析与风险评估	foundation model multimodal
3	Evaluating the Reliability and Fidelity of Automated Judgment Systems of Large Language Models	评估大型语言模型作为自动评估系统的可靠性和保真度	large language model
4	MARCUS: An agentic, multimodal vision-language model for cardiac diagnosis and management	MARCUS：用于心脏诊断和管理的Agentic多模态视觉-语言模型	multimodal
5	Reasoning or Rhetoric? An Empirical Analysis of Moral Reasoning Explanations in Large Language Models	分析大型语言模型在道德推理中是否仅为修辞，揭示其与人类道德发展的不一致性。	large language model
6	AI Token Futures Market: Commoditization of Compute and Derivatives Contract Design	提出AI Token期货市场，实现算力商品化及衍生品合约设计	vision-language-action VLA large language model
7	Deterministic Hallucination Detection in Medical VQA via Confidence-Evidence Bayesian Gain	提出CEBaG，一种确定性的医学VQA幻觉检测方法，无需采样和外部模型。	large language model multimodal
8	Compensating Visual Insufficiency with Stratified Language Guidance for Long-Tail Class Incremental Learning	提出分层语言引导方法，解决长尾类增量学习中的视觉信息不足问题	large language model
9	SecureBreak -- A dataset towards safe and secure models	提出SecureBreak数据集，用于提升大型语言模型安全性与防御对抗攻击能力	large language model
10	CurvZO: Adaptive Curvature-Guided Sparse Zeroth-Order Optimization for Efficient LLM Fine-Tuning	CurvZO：自适应曲率引导的稀疏零阶优化，用于高效LLM微调	large language model
11	Cognitive Agency Surrender: Defending Epistemic Sovereignty via Scaffolded AI Friction	提出脚手架式认知摩擦，防御认知代理权让渡，保障认知主权。	multimodal
12	Silicon Bureaucracy and AI Test-Oriented Education: Contamination Sensitivity and Score Confidence in LLM Benchmarks	提出LLM基准测试污染敏感性和置信度审计框架，评估基准测试的可靠性。	large language model
13	AgenticRec: End-to-End Tool-Integrated Policy Optimization for Ranking-Oriented Recommender Agents	AgenticRec：面向排序的推荐Agent端到端工具集成策略优化	large language model
14	LLM-Based Test Case Generation in DBMS through Monte Carlo Tree Search	MIST：基于蒙特卡洛树搜索的LLM驱动DBMS测试用例生成框架	large language model
15	Unified-MAS: Universally Generating Domain-Specific Nodes for Empowering Automatic Multi-Agent Systems	Unified-MAS：通过通用领域节点生成增强自动多智能体系统	chain-of-thought

🔬 支柱二：RL算法与架构 (RL & Architecture) (9 篇)

#	题目	一句话要点	标签	🔗	⭐
16	SpecTM: Spectral Targeted Masking for Trustworthy Foundation Models	SpecTM：面向可信基础模型的谱段针对性掩码策略	predictive model representation learning foundation model
17	Suiren-1.0 Technical Report: A Family of Molecular Foundation Models	Suiren-1.0：构建分子领域基础模型，实现量子性质预测与高效下游应用	distillation foundation model
18	EvoIdeator: Evolving Scientific Ideas through Checklist-Grounded Reinforcement Learning	提出EvoIdeator以解决科学创意生成中的反馈不足问题	reinforcement learning large language model
19	Spatio-Temporal Attention Enhanced Multi-Agent DRL for UAV-Assisted Wireless Networks with Limited Communications	针对通信受限的无人机网络，提出时空注意力增强的多智能体DRL算法	reinforcement learning deep reinforcement learning DRL
20	Adaptive Robust Estimator for Multi-Agent Reinforcement Learning	提出DACR和ARE框架，解决多智能体强化学习中的信用分配和噪声奖励问题	reinforcement learning large language model
21	DiT-Flow: Speech Enhancement Robust to Multiple Distortions based on Flow Matching in Latent Space and Diffusion Transformers	提出基于Flow Matching和Diffusion Transformer的DiT-Flow，提升多重失真下的语音增强鲁棒性。	flow matching
22	Counterfactual Credit Policy Optimization for Multi-Agent Collaboration	提出CCPO，通过反事实推理优化多智能体LLM协作中的信用分配问题	reinforcement learning large language model	✅
23	A Context Engineering Framework for Improving Enterprise AI Agents based on Digital-Twin MDP	提出基于数字孪生MDP的上下文工程框架，提升企业AI Agent性能	reinforcement learning offline reinforcement learning
24	RuntimeSlicer: Towards Generalizable Unified Runtime State Representation for Failure Management	RuntimeSlicer：面向可泛化的统一运行时状态表示，用于故障管理	representation learning contrastive learning

🔬 支柱六：视频提取与匹配 (Video Extraction) (1 篇)

#	题目	一句话要点	标签	🔗	⭐
25	Mind over Space: Can Multimodal Large Language Models Mentally Navigate?	提出NavMind模型，提升多模态大语言模型在复杂环境下的心智导航能力	egocentric spatiotemporal large language model

🔬 支柱五：交互与反应 (Interaction & Reaction) (1 篇)

#	题目	一句话要点	标签	🔗	⭐
26	Towards Secure Retrieval-Augmented Generation: A Comprehensive Review of Threats, Defenses and Benchmarks	全面剖析RAG安全：威胁、防御与基准评测，保障可信赖的知识增强生成。	OMOMO large language model

⬅️ 返回 cs.AI 首页 · 🏠 返回主页