cs.LG（2025-04-25）

📊 共 14 篇论文 | 🔗 1 篇有代码

🎯 兴趣领域导航

支柱二：RL算法与架构 (RL & Architecture) (8 🔗1) 支柱九：具身大模型 (Embodied Foundation Models) (5) 支柱三：空间感知与语义 (Perception & Semantics) (1)

🔬 支柱二：RL算法与架构 (RL & Architecture) (8 篇)

#题目一句话要点标签🔗⭐
1 Multimodal graph representation learning for website generation based on visual sketch 提出基于多模态图表示学习的网站生成方法，提升设计到代码的自动化水平。 representation learning multimodal ✅
2 Deep Reinforcement Learning Based Navigation with Macro Actions and Topological Maps 提出基于宏动作和拓扑地图的深度强化学习导航方法，解决复杂环境下的稀疏奖励问题。 reinforcement learning deep reinforcement learning
3 Generalization Capability for Imitation Learning 基于信息论与数据分布特性的模仿学习泛化能力分析框架 imitation learning foundation model
4 Foundations of Safe Online Reinforcement Learning in the Linear Quadratic Regulator: $\sqrt{T}$-Regret 针对安全约束线性二次调节器，提出$\tilde{O}_T(\sqrt{T})$遗憾度的安全在线强化学习算法 reinforcement learning
5 Representation Learning for Distributional Perturbation Extrapolation 提出扰动分布自编码器(PDAE)，用于预测未见扰动下的测量分布。 representation learning
6 Generalization Guarantees for Multi-View Representation Learning and Application to Regularization via Gaussian Product Mixture Prior 针对多视角表征学习，提出基于高斯乘积混合先验的正则化方法，提升泛化性能。 representation learning
7 A Generative Graph Contrastive Learning Model with Global Signal 提出CSG2L框架，通过全局信号生成对比学习样本，提升图学习性能。 contrastive learning
8 Learning from Less: SINDy Surrogates in RL 提出基于SINDy的强化学习代理环境，降低计算成本并保持性能。 reinforcement learning model-based RL

🔬 支柱九：具身大模型 (Embodied Foundation Models) (5 篇)

#题目一句话要点标签🔗⭐
9 TLoRA: Tri-Matrix Low-Rank Adaptation of Large Language Models TLoRA：一种用于大型语言模型的三矩阵低秩自适应方法 large language model
10 A Unified MDL-based Binning and Tensor Factorization Framework for PDF Estimation 提出基于MDL的联合分箱与张量分解框架，用于概率密度函数估计 multimodal
11 NoEsis: Differentially Private Knowledge Transfer in Modular LLM Adaptation NoEsis：模块化LLM自适应中的差分隐私知识迁移框架 large language model
12 Think, Prune, Train, Improve: Scaling Reasoning without Scaling Models 提出Think, Prune, Train框架，无需扩大模型即可提升LLM推理能力 large language model
13 Streaming, Fast and Slow: Cognitive Load-Aware Streaming for Efficient LLM Serving 提出认知负荷感知的LLM流式服务，提升效率并节约计算资源 large language model

🔬 支柱三：空间感知与语义 (Perception & Semantics) (1 篇)

#题目一句话要点标签🔗⭐
14 Three Types of Calibration with Properties and their Semantic and Formal Relationships 提出基于属性的三种校准方法，并分析其语义和形式关系，以应对预测系统校准概念碎片化问题。 semantic map

⬅️ 返回 cs.LG 首页 · 🏠 返回主页