cs.LG（2025-03-31）

📊 共 25 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) (11 🔗3) 支柱二：RL算法与架构 (RL & Architecture) (9 🔗1) 支柱一：机器人控制 (Robot Control) (2) 支柱四：生成式动作 (Generative Motion) (2) 支柱八：物理动画 (Physics-based Animation) (1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (11 篇)

#题目一句话要点标签🔗⭐
1 ORAL: Prompting Your Large-Scale LoRAs via Conditional Recurrent Diffusion ORAL：通过条件循环扩散模型提示大规模LoRA，实现可控且可扩展的参数生成。 large language model foundation model multimodal
2 Predicting Targeted Therapy Resistance in Non-Small Cell Lung Cancer Using Multimodal Machine Learning 提出一种多模态机器学习模型，用于预测非小细胞肺癌患者对奥希替尼的耐药性。 multimodal
3 LLM4FS: Leveraging Large Language Models for Feature Selection LLM4FS：利用大语言模型进行特征选择的混合策略 large language model ✅
4 Rethinking Key-Value Cache Compression Techniques for Large Language Model Serving 重新审视大语言模型服务的键值缓存压缩技术，提升实际部署性能 large language model ✅
5 Translating Multimodal AI into Real-World Inspection: TEMAI Evaluation Framework and Pathways for Implementation 提出TEMAI框架，评估多模态AI在工业检测中的转化能力与实施路径 multimodal
6 Communication-Efficient and Personalized Federated Foundation Model Fine-Tuning via Tri-Matrix Adaptation 提出CE-LoRA，通过三矩阵适应实现通信高效的个性化联邦大模型微调 foundation model
7 Timeseries Foundation Models for Mobility: A Benchmark Comparison with Traditional and Deep Learning Models 评估时间序列预训练模型在城市出行预测中的性能，对比传统与深度学习方法。 foundation model
8 Effectively Controlling Reasoning Models through Thinking Intervention 提出思维干预方法，有效控制推理型大语言模型的推理过程 large language model instruction following
9 Inference-Time Scaling for Complex Tasks: Where We Stand and What Lies Ahead 研究推理时扩展对复杂任务的影响，揭示其局限性与未来潜力 large language model
10 Evaluating and Designing Sparse Autoencoders by Approximating Quasi-Orthogonality 提出基于近似准正交性的稀疏自编码器评估与设计方法，解决超参数k选择难题。 large language model ✅
11 Green MLOps to Green GenOps: An Empirical Study of Energy Consumption in Discriminative and Generative AI Operations 研究判别式与生成式AI模型在MLOps流程中的能耗，为绿色GenOps提供实践指导。 large language model

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

#题目一句话要点标签🔗⭐
12 Unimodal-driven Distillation in Multimodal Emotion Recognition with Dynamic Fusion 提出SUMMER框架，利用单模态知识蒸馏提升多模态情感识别性能 distillation multimodal
13 TransMamba: A Sequence-Level Hybrid Transformer-Mamba Language Model 提出TransMamba，一种序列级混合Transformer-Mamba语言模型，提升长序列建模效率。 Mamba SSM state space model ✅
14 Scalable Ride-Sourcing Vehicle Rebalancing with Service Accessibility Guarantee: A Constrained Mean-Field Reinforcement Learning Approach 提出基于约束均值场强化学习的可扩展网约车再平衡方法，保障服务可达性 reinforcement learning spatiotemporal
15 Open-Reasoner-Zero: An Open Source Approach to Scaling Up Reinforcement Learning on the Base Model 开源Open-Reasoner-Zero：在基础模型上扩展强化学习，提升推理能力 reinforcement learning PPO
16 Node Embeddings via Neighbor Embeddings 提出图邻居嵌入(graph NE)框架，直接聚合相邻节点嵌入向量，提升局部结构保持能力。 representation learning contrastive learning structure preservation
17 Fair Dynamic Spectrum Access via Fully Decentralized Multi-Agent Reinforcement Learning 提出基于多智能体强化学习的公平动态频谱接入方案，实现去中心化网络资源优化。 reinforcement learning
18 Level the Level: Balancing Game Levels for Asymmetric Player Archetypes With Reinforcement Learning 提出基于强化学习的关卡平衡方法，解决非对称多人游戏中关卡设计难题 reinforcement learning
19 Accelerating High-Efficiency Organic Photovoltaic Discovery via Pretrained Graph Neural Networks and Generative Reinforcement Learning 利用预训练图神经网络和生成式强化学习加速高效有机光伏材料的发现 reinforcement learning
20 Dynamic Operating System Scheduling Using Double DQN: A Reinforcement Learning Approach to Task Optimization 提出基于双重DQN的动态调度算法以优化任务调度 reinforcement learning

🔬 支柱一：机器人控制 (Robot Control) (2 篇)

#题目一句话要点标签🔗⭐
21 A Survey of Reinforcement Learning-Based Motion Planning for Autonomous Driving: Lessons Learned from a Driving Task Perspective 综述基于强化学习的自动驾驶运动规划，从驾驶任务视角提炼经验与挑战。 motion planning reinforcement learning
22 An extension of linear self-attention for in-context learning 扩展线性自注意力机制，提升Transformer上下文学习能力 manipulation

🔬 支柱四：生成式动作 (Generative Motion) (2 篇)

#题目一句话要点标签🔗⭐
23 DiffScale: Continuous Downscaling and Bias Correction of Subseasonal Wind Speed Forecasts using Diffusion Models DiffScale：利用扩散模型连续降尺度和偏差校正次季节风速预测 classifier-free guidance
24 Physics-informed neural networks for hidden boundary detection and flow field reconstruction 提出基于物理信息的神经网络，用于流场中隐藏边界检测与流场重建 penetration

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

#题目一句话要点标签🔗⭐
25 Data-Driven Forecasting of High-Dimensional Transient and Stationary Processes via Space-Time Projection 提出时空投影(STP)方法，用于高维瞬态和稳态过程的数据驱动预测。 spatiotemporal

⬅️ 返回 cs.LG 首页 · 🏠 返回主页