cs.AI（2026-03-10）

📊 共 23 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) (17 🔗1) 支柱二：RL算法与架构 (RL & Architecture) (4) 支柱八：物理动画 (Physics-based Animation) (2)

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

#	题目	一句话要点	标签	🔗	⭐
1	MedMASLab: A Unified Orchestration Framework for Benchmarking Multimodal Medical Multi-Agent Systems	MedMASLab：用于多模态医学多智能体系统基准测试的统一编排框架	multimodal visual grounding	✅
2	Quantifying the Necessity of Chain of Thought through Opaque Serial Depth	通过不透明串行深度量化思维链的必要性	large language model chain-of-thought
3	World2Mind: Cognition Toolkit for Allocentric Spatial Reasoning in Foundation Models	World2Mind：用于具身智能体认知空间推理的工具包	foundation model multimodal
4	GenePlan: Evolving Better Generalized PDDL Plans using Large Language Models	GenePlan：利用大语言模型进化生成更优的通用PDDL规划器	large language model chain-of-thought
5	EsoLang-Bench: Evaluating Genuine Reasoning in Large Language Models via Esoteric Programming Languages	EsoLang-Bench：通过冷门编程语言评估大语言模型的真正推理能力	large language model
6	Understanding the Use of a Large Language Model-Powered Guide to Make Virtual Reality Accessible for Blind and Low Vision People	利用大型语言模型为视障人士打造可交互的虚拟现实辅助工具	large language model
7	PathMem: Toward Cognition-Aligned Memory Transformation for Pathology MLLMs	提出PathMem以解决病理MLLMs知识整合不足问题	large language model multimodal
8	Think Before You Lie: How Reasoning Improves Honesty	推理提升大语言模型诚实度：揭示表征空间几何与道德决策的关系	large language model
9	The Reasoning Trap -- Logical Reasoning as a Mechanistic Pathway to Situational Awareness	提出RAISE框架，揭示逻辑推理能力提升如何驱动AI系统涌现情境感知能力	large language model
10	MITRA: An AI Assistant for Knowledge Retrieval in Physics Collaborations	MITRA：用于物理合作中知识检索的AI助手，解决信息爆炸难题	large language model
11	Does the Question Really Matter? Training-Free Data Selection for Vision-Language SFT	提出CVS：一种免训练的视觉-语言SFT数据选择方法，提升多模态推理能力。	multimodal
12	MUGEN: Evaluating and Improving Multi-audio Understanding of Large Audio-Language Models	MUGEN：评估并提升大语音语言模型的多语音理解能力	chain-of-thought
13	MiniAppBench: Evaluating the Shift from Text to Interactive HTML Responses in LLM-Powered Assistants	提出MiniAppBench，评估LLM驱动助手中从文本到交互式HTML响应的转变	large language model
14	Curveball Steering: The Right Direction To Steer Isn't Always Linear	提出Curveball Steering，通过非线性干预提升大语言模型行为控制效果	large language model
15	Cognitively Layered Data Synthesis for Domain Adaptation of LLMs to Space Situational Awareness	提出BD-FDG框架，用于领域自适应LLM在空间态势感知中的应用	large language model
16	Real-Time Trust Verification for Safe Agentic Actions using TrustBench	TrustBench：用于Agent安全行动的实时信任验证框架	large language model
17	Deep Tabular Research via Continual Experience-Driven Execution	提出基于持续经验驱动执行的深度表格研究框架，解决复杂表格推理难题	large language model

🔬 支柱二：RL算法与架构 (RL & Architecture) (4 篇)

#	题目	一句话要点	标签	🔗	⭐
18	OOD-MMSafe: Advancing MLLM Safety from Harmful Intent to Hidden Consequences	提出OOD-MMSafe基准与CASPO框架，提升MLLM在复杂因果链中的安全推理能力	distillation large language model multimodal
19	Robust Regularized Policy Iteration under Transition Uncertainty	提出RRPI算法，通过鲁棒正则化策略迭代解决离线强化学习中的转移不确定性问题	reinforcement learning policy learning offline RL
20	Influencing LLM Multi-Agent Dialogue via Policy-Parameterized Prompts	提出基于策略参数化Prompt的LLM多智能体对话控制方法	reinforcement learning large language model
21	Social-R1: Towards Human-like Social Reasoning in LLMs	提出Social-R1框架，提升LLM在复杂社交推理中的类人能力	reinforcement learning large language model

🔬 支柱八：物理动画 (Physics-based Animation) (2 篇)

#	题目	一句话要点	标签	🔗	⭐
22	Physics-Informed Neural Engine Sound Modeling with Differentiable Pulse-Train Synthesis	提出基于物理信息的神经引擎声音建模方法，实现可微分的脉冲序列合成。	PULSE
23	A Variational Latent Equilibrium for Learning in Cortex	提出基于变分隐平衡的皮层学习模型，逼近生物可解释的时序反向传播	spatiotemporal

⬅️ 返回 cs.AI 首页 · 🏠 返回主页