cs.AI（2025-05-19）

📊 共 24 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) (15 🔗3) 支柱二：RL算法与架构 (RL & Architecture) (6 🔗1) 支柱一：机器人控制 (Robot Control) (2) 支柱八：物理动画 (Physics-based Animation) (1)

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

#题目一句话要点标签🔗⭐
1 MM-PRM: Enhancing Multimodal Mathematical Reasoning with Scalable Step-Level Supervision 提出MM-PRM，通过可扩展的步骤级监督增强多模态数学推理能力 large language model multimodal ✅
2 Aneumo: A Large-Scale Multimodal Aneurysm Dataset with Computational Fluid Dynamics Simulations and Deep Learning Benchmarks Aneumo：大规模多模态脑动脉瘤数据集，含CFD模拟与深度学习基准。 multimodal ✅
3 Survey: Multi-Armed Bandits Meet Large Language Models 探索多臂老虎机与大语言模型的协同：优化决策与自然语言处理 large language model
4 Advancing Software Quality: A Standards-Focused Review of LLM-Based Assurance Techniques 利用LLM提升软件质量：一项聚焦标准的软件质量保证技术综述 large language model multimodal
5 Causal Head Gating: A Framework for Interpreting Roles of Attention Heads in Transformers 提出因果头门控（CHG）框架，用于Transformer模型中注意力头的功能角色解释。 large language model instruction following
6 Unified Cross-modal Translation of Score Images, Symbolic Music, and Performance Audio 提出统一跨模态翻译框架，实现乐谱图像、符号音乐和演奏音频间的转换。 multimodal
7 Ice Cream Doesn't Cause Drowning: Benchmarking LLMs Against Statistical Pitfalls in Causal Inference CausalPitfalls：评估LLM在因果推断中应对统计陷阱能力的基准测试 large language model
8 Security Degradation in Iterative AI Code Generation -- A Systematic Analysis of the Paradox 揭示AI代码迭代生成中的安全退化悖论，强调人工干预的重要性 large language model
9 Safety Alignment Can Be Not Superficial With Explicit Safety Signals 引入显式安全信号，提升大语言模型对抗攻击的鲁棒性 large language model
10 Measuring the Faithfulness of Thinking Drafts in Large Reasoning Models 提出反事实干预框架，评估大型推理模型中思维草稿的忠实性 chain-of-thought
11 Language Models Are Capable of Metacognitive Monitoring and Control of Their Internal Activations 提出神经反馈范式以量化语言模型的元认知能力 large language model
12 CoT-Kinetics: A Theoretical Modeling Assessing LRM Reasoning Process 提出CoT-Kinetics能量方程，评估大型推理模型(LRM)推理过程的合理性。 large language model
13 AutoMathKG: The automated mathematical knowledge graph based on LLM and vector database AutoMathKG：提出一种基于LLM和向量数据库的自动化数学知识图谱构建方法 large language model
14 CompeteSMoE -- Statistically Guaranteed Mixture of Experts Training via Competition 提出CompeteSMoE以解决稀疏专家模型训练效率问题 large language model ✅
15 LLM-KG-Bench 3.0: A Compass for SemanticTechnology Capabilities in the Ocean of LLMs LLM-KG-Bench 3.0：评估大语言模型在语义技术和知识图谱工程能力的基准框架 large language model

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

#题目一句话要点标签🔗⭐
16 Trust, But Verify: A Self-Verification Approach to Reinforcement Learning with Verifiable Rewards 提出RISE框架，通过在线自验证强化学习提升LLM数学推理能力 reinforcement learning large language model
17 Warm Up Before You Train: Unlocking General Reasoning in Resource-Constrained Settings 提出Warm Up预训练策略，提升资源受限场景下LLM的通用推理能力 reinforcement learning distillation chain-of-thought
18 CoIn: Counting the Invisible Reasoning Tokens in Commercial Opaque LLM APIs CoIn：审计商业闭源LLM API中不可见的推理token，解决token数量欺诈问题 reinforcement learning large language model ✅
19 Building spatial world models from sparse transitional episodic memories 提出Episodic Spatial World Model，从稀疏情景记忆构建空间世界模型 world model
20 Exploiting Symbolic Heuristics for the Synthesis of Domain-Specific Temporal Planning Guidance using Reinforcement Learning 利用符号启发式，通过强化学习合成领域特定时序规划指导 reinforcement learning
21 QUADS: QUAntized Distillation Framework for Efficient Speech Language Understanding 提出QUADS框架，用于高效语音语言理解的量化蒸馏 distillation

🔬 支柱一：机器人控制 (Robot Control) (2 篇)

#题目一句话要点标签🔗⭐
22 Large Language Models and Their Applications in Roadway Safety and Mobility Enhancement: A Comprehensive Review 综述：大型语言模型在道路安全与交通优化中的应用 sim-to-real large language model foundation model
23 Scaling Computer-Use Grounding via User Interface Decomposition and Synthesis 提出OSWorld-G基准和Jedi数据集，提升计算机使用Agent的GUI Grounding能力。 manipulation foundation model

🔬 支柱八：物理动画 (Physics-based Animation) (1 篇)

#题目一句话要点标签🔗⭐
24 Direction-Aware Neural Acoustic Fields for Few-Shot Interpolation of Ambisonic Impulse Responses 提出方向感知神经声场，用于少样本混响脉冲响应插值 PULSE

⬅️ 返回 cs.AI 首页 · 🏠 返回主页