cs.AI（2026-06-09）

📊 共 39 篇论文 | 🔗 6 篇有代码

🎯 兴趣领域导航

支柱九：具身大模型 (Embodied Foundation Models) (25 🔗3) 支柱二：RL算法与架构 (RL & Architecture) (12 🔗2) 支柱三：空间感知与语义 (Perception & Semantics) (2 🔗1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (25 篇)

#题目一句话要点标签🔗⭐
1 When the Chain of Thought Knows Better: Failure Modes in Multi-Turn Reasoning Models 提出CoT-Output安全矩阵以揭示多轮推理模型的失败模式 chain-of-thought
2 Null-Space Constrained Low-Rank Adaptation for Response-Specified Large Language Model Unlearning 提出Null-Space约束低秩适应以解决大语言模型的遗忘问题 large language model
3 A History-Aware Visually Grounded Critic for Computer Use Agents 提出HiViG框架以解决CUA决策短视与视觉缺失问题 multimodal visual grounding
4 ReasonAlloc: Hierarchical Decoding-Time KV Cache Budget Allocation for Reasoning Models 提出ReasonAlloc以解决推理模型中的KV缓存预算分配问题 large language model chain-of-thought
5 Towards Autonomous Accelerator Design: FPGA Accelerator Generation with SECDA 提出SECDA-DSE框架以自动化FPGA加速器设计 large language model chain-of-thought
6 Soul Computing: A Theoretical Framework and Technical Architecture for Intelligent Agents with Independent Consciousness 提出灵魂计算框架以解决智能体独立意识构建问题 large language model multimodal
7 Generative Explainability for Next-Generation Networks: LLM-Augmented XAI with Mutual Feature Interactions 提出生成性可解释性框架以解决网络透明性问题 large language model
8 A Constrained Natural-Language Interface for Variational Multi-Physics Finite Element Simulations in FEniCS 提出受限自然语言接口以优化多物理场有限元模拟 large language model
9 Piper: A Programmable Distributed Training System 提出Piper以解决大规模模型训练的并行策略适应性问题 foundation model
10 Flaws in the LLM Automation Narrative 提出新基准任务以评估LLM在数据分析中的表现 large language model
11 Superficial Beliefs in LLM Decision-Making 提出对LLM决策中的表面信念的分析以揭示决策结构 large language model
12 Structure from Reasoning, Numbers from Search: On-Premise Open LLMs as Structural Priors for Coupled MIMO Controller Tuning 利用开放源代码大语言模型优化耦合MIMO控制器调优 large language model
13 Mind the Gap: Can Frontier LLMs Pass a Standardized Office Proficiency Exam? 提出基于NCRE的评估方法以测试前沿LLM在办公自动化中的能力 large language model
14 Role-Agent: Bootstrapping LLM Agents via Dual-Role Evolution 提出Role-Agent框架以解决LLM代理学习中的反馈效率问题 large language model
15 Evaluating Research-Level Math Proofs via Strict Step-Level Verification 提出严格逐步验证框架以解决数学证明评估问题 large language model
16 Toward Secure LLM Agents: Threat Surfaces, Attacks, Defenses, and Evaluation 提出安全LLM代理的综合框架以应对新兴威胁 large language model
17 Decentralized Multi-Agent Systems with Shared Context 提出去中心化语言模型以解决多智能体系统的协调瓶颈问题 large language model ✅
18 ActiveMem: Distributed Active Memory for Long-Horizon LLM Reasoning 提出ActiveMem以解决长时间推理任务中的记忆管理问题 large language model
19 Decoupling Thought from Speech: Knowledge-Grounded Counterfactual Reasoning for Resilient Multi-Agent Argumentation 提出知识基础的反事实推理以解决多智能体辩论中的稳定性问题 large language model
20 STAGE-Claw: Automated State-based Agent Benchmarking for Realistic Scenarios 提出STAGE-Claw框架以解决个人代理评估的挑战 large language model
21 Beyond Static Evaluation: Co-Evolutionary Mechanisms for LLM-Driven Strategy Evolution in Adversarial Games 提出共进化机制以解决对抗游戏中策略演化的评估挑战 foundation model ✅
22 Atomic Intent Reasoning: Bringing LLM Semantics to Industrial Cross-Domain Recommendations 提出AIR框架以解决跨域推荐中的语义差距问题 large language model
23 Mobility Anomaly Generation using LLM-Driven Behavior with Kinematic Constraints 提出基于LLM的运动异常生成框架以解决数据稀缺问题 large language model
24 From Context-Aware to Conflict-Aware: Generalizing Contrastive Decoding for Knowledge Conflict in LLMs 提出冲突感知解码方法以解决大语言模型中的知识冲突问题 large language model ✅
25 Sim2Schedule: A Simulator-Guided LLM Framework for Autonomous Open-Pit Mine Scheduling 提出Sim2Schedule框架以解决开放式矿山调度问题 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🔗⭐
26 One Token per Multimodal Evidence: Latent Memory for Resource-Constrained QA 提出潜在记忆以解决资源受限的多模态问答问题 distillation large language model multimodal ✅
27 ReflectiChain: Epistemic Grounding in LLM-Driven World Models for Supply Chain Resilience 提出REFLECTICHAIN以解决供应链中的认知差距问题 reinforcement learning world model world models
28 WorldKernel: A World Model is the Coupling Kernel of Admissible Possible Worlds 提出WorldKernel模型以解决反事实世界耦合预测问题 world model world models
29 Speech Meets ELF: Audio Conditional Continuous-Target Diffusion for Speech Recognition and Translation 提出ELF-S2T以解决语音识别与翻译中的离散生成问题 flow matching semantic mapping semantic map ✅
30 Cross-Modal Knowledge Distillation without Paired Data: Theoretical Foundation and Algorithm 提出无配对数据的跨模态知识蒸馏方法以解决数据获取难题 distillation multimodal
31 Reasoning or Memorization? Direction-Aware Diversity Exploration in LLM Reinforcement Learning 提出DiRL框架以解决LLM探索中的推理与记忆问题 reinforcement learning large language model
32 Self-Distillation Policy Optimization via Visual Feedback: Bridging Code and Visual Artifacts 提出视觉反馈自蒸馏策略优化框架以解决代码生成视觉缺陷问题 distillation large language model
33 Dmsh: A Multi-Agent Reinforcement Learning Framework for All-Quad Mesh Generation 提出Dmsh框架以解决全四边形网格生成问题 reinforcement learning curriculum learning
34 Bellman-Taylor Score Decoding for Markov Decision Processes with State-Dependent Feasible Action Sets 提出Bellman-Taylor评分解码以解决状态依赖的MDP问题 reinforcement learning deep reinforcement learning DRL
35 The Role of Feedback Alignment in Self-Distillation 提出自蒸馏中的反馈对齐方法以提升模型性能 distillation
36 Effective Reinforcement Learning for Agentic Search by Recycling Zero-Variance Queries During Training 提出查询回收机制以提升强化学习搜索代理的训练效率 reinforcement learning
37 HIPIF: Hierarchical Planning and Information Folding for Long-Horizon LLM Agent Learning 提出HIPIF以解决长时间跨度LLM代理学习中的上下文干扰问题 reinforcement learning large language model

🔬 支柱三：空间感知与语义 (Perception & Semantics) (2 篇)

#题目一句话要点标签🔗⭐
38 Spatial-Omni: Spatial Audio Understanding Integration in Multimodal LLMs via FOA Encoding 提出Spatial-Omni以解决空间音频理解不足问题 scene understanding large language model multimodal ✅
39 Large-scale semantic mapping of learner agency and autonomy reveals what measurement and generative AI research overlook 提出语义映射方法以解决学习者自主性测量不足问题 semantic mapping semantic map

⬅️ 返回 cs.AI 首页 · 🏠 返回主页