cs.AI（2026-03-09）

📊 共 20 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) (14 🔗1) 支柱二：RL算法与架构 (RL & Architecture) (6)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (14 篇)

#	题目	一句话要点	标签	🔗	⭐
1	CMMR-VLN: Vision-and-Language Navigation via Continual Multimodal Memory Retrieval	CMMR-VLN：提出基于持续多模态记忆检索的视觉语言导航框架，提升长程和未知环境下的导航性能。	VLN large language model multimodal
2	M$^3$-ACE: Rectifying Visual Perception in Multimodal Math Reasoning via Multi-Agentic Context Engineering	提出M$^3$-ACE框架，通过多智能体上下文工程提升多模态数学推理中的视觉感知准确性。	large language model multimodal
3	Evaluating Financial Intelligence in Large Language Models: Benchmarking SuperInvesting AI with LLM Engines	提出AI金融智能基准AFIB，评估大语言模型在金融分析中的能力。	large language model
4	Deconstructing Multimodal Mathematical Reasoning: Towards a Unified Perception-Alignment-Reasoning Paradigm	解构多模态数学推理，提出统一的感知-对齐-推理范式	multimodal
5	CoCo: Code as CoT for Text-to-Image Preview and Rare Concept Generation	提出CoCo：一种基于代码的CoT框架，用于文本到图像的预览和罕见概念生成。	multimodal chain-of-thought	✅
6	CORE-Acu: Structured Reasoning Traces and Knowledge Graph Safety Verification for Acupuncture Clinical Decision Support	提出CORE-Acu框架，结合结构化推理和知识图谱安全验证，提升针灸临床决策支持的可靠性。	large language model chain-of-thought
7	PIRA-Bench: A Transition from Reactive GUI Agents to GUI-based Proactive Intent Recommendation Agents	提出PIRA-Bench基准，用于评估GUI环境下主动意图推荐Agent	large language model multimodal
8	AI Agents, Language, Deep Learning and the Next Revolution in Science	提出基于大语言模型的AI Agent，赋能科研数据分析与知识发现	large language model multimodal
9	The Struggle Between Continuation and Refusal: A Mechanistic Analysis of the Continuation-Triggered Jailbreak in LLMs	针对LLM中延续触发的越狱现象，提出基于注意力头的机制性分析方法	large language model
10	CDRRM: Contrast-Driven Rubric Generation for Reliable and Interpretable Reward Modeling	提出CDRRM，通过对比驱动的准则生成，实现可靠且可解释的奖励建模。	large language model
11	Agentic Neurosymbolic Collaboration for Mathematical Discovery: A Case Study in Combinatorial Design	通过神经符号协作实现组合设计中的数学发现	large language model
12	Adaptive Collaboration with Humans: Metacognitive Policy Optimization for Multi-Agent LLMs with Continual Learning	提出人机协同多智能体框架HILA，解决多智能体LLM在开放世界中的知识局限性问题。	large language model
13	SWE-Fuse: Empowering Software Agents via Issue-free Trajectory Learning and Entropy-aware RLVR Training	SWE-Fuse：通过无问题轨迹学习和熵感知RLVR训练增强软件智能体	large language model
14	Ares: Adaptive Reasoning Effort Selection for Efficient LLM Agents	Ares：自适应推理努力选择框架，提升LLM Agent效率	chain-of-thought

🔬 支柱二：RL算法与架构 (RL & Architecture) (6 篇)

#	题目	一句话要点	标签	🔗	⭐
15	Disentangling Reasoning in Large Audio-Language Models for Ambiguous Emotion Prediction	提出一种基于大语音语言模型（LALM）的解耦推理框架，用于解决语音情感识别中的歧义性情感预测问题。	DPO motion prediction chain-of-thought
16	In-Context Reinforcement Learning for Tool Use in Large Language Models	提出ICRL，一种无需SFT的上下文强化学习方法，提升LLM工具使用能力	reinforcement learning large language model
17	Agentic Critical Training	提出Agentic Critical Training，提升LLM智能体自主推理和行动质量评估能力	reinforcement learning imitation learning distillation
18	The Boiling Frog Threshold: Criticality and Blindness in World Model-Based Anomaly Detection Under Gradual Drift	研究基于世界模型的强化学习智能体在渐进漂移下的异常检测阈值与盲区问题	world model
19	RetroAgent: From Solving to Evolving via Retrospective Dual Intrinsic Feedback	RetroAgent：通过回顾式双重内在反馈，实现LLM智能体从问题解决到持续进化	reinforcement learning large language model
20	Efficient Policy Learning with Hybrid Evaluation-Based Genetic Programming for Uncertain Agile Earth Observation Satellite Scheduling	提出混合评估遗传编程(HE-GP)算法，高效解决不确定敏捷地球观测卫星调度问题。	policy learning

⬅️ 返回 cs.AI 首页 · 🏠 返回主页