cs.LG（2026-03-25）

📊 共 23 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) (14 🔗1) 支柱九：具身大模型 (Embodied Foundation Models) (7) 支柱一：机器人控制 (Robot Control) (1) 支柱八：物理动画 (Physics-based Animation) (1)

🔬 支柱二：RL算法与架构 (RL & Architecture) (14 篇)

#题目一句话要点标签🔗⭐
1 DreamerAD: Efficient Reinforcement Learning via Latent World Model for Autonomous Driving DreamerAD：基于潜空间世界模型的自动驾驶高效强化学习 reinforcement learning world model world models
2 Large Language Model Guided Incentive Aware Reward Design for Cooperative Multi-Agent Reinforcement Learning 提出基于大语言模型的激励感知奖励设计，用于合作多智能体强化学习 reinforcement learning reward design large language model
3 Can we generate portable representations for clinical time series data using LLMs? 利用LLM生成临床时间序列数据的可迁移表征，提升模型泛化能力 predictive model representation learning large language model
4 HDPO: Hybrid Distillation Policy Optimization via Privileged Self-Distillation 提出HDPO，通过特权自蒸馏解决数学推理中强化学习的“悬崖”问题。 reinforcement learning distillation large language model
5 UI-Voyager: A Self-Evolving GUI Agent Learning via Failed Experience UI-Voyager：提出一种基于失败经验自进化的GUI智能体，提升移动GUI自动化性能 distillation large language model multimodal
6 TuneShift-KD: Knowledge Distillation and Transfer for Fine-tuned Models 提出TuneShift-KD，通过困惑度差异蒸馏微调模型中的领域知识到新模型。 distillation foundation model
7 CUA-Suite: Massive Human-annotated Video Demonstrations for Computer-Use Agents CUA-Suite：大规模人工标注视频数据集，助力计算机使用智能体研究 world model world models multimodal
8 Marchuk: Efficient Global Weather Forecasting from Mid-Range to Sub-Seasonal Scales via Flow Matching Marchuk：基于流匹配的高效全球中长期天气预测模型 flow matching ✅
9 CGRL: Causal-Guided Representation Learning for Graph Out-of-Distribution Generalization 提出CGRL，通过因果引导表示学习提升图神经网络的OOD泛化能力 representation learning
10 A Deep Dive into Scaling RL for Code Generation with Synthetic Data and Curricula 提出基于多轮合成数据和课程学习的强化学习方法，提升代码生成能力。 reinforcement learning large language model
11 Causality-Driven Disentangled Representation Learning in Multiplex Graphs 提出CaDeM框架，通过因果推断解耦多重图中的共享与私有表示 representation learning
12 KCLNet: Electrically Equivalence-Oriented Graph Representation Learning for Analog Circuits 提出KCLNet，用于模拟电路的电气等效性导向图表示学习 representation learning
13 Towards Effective Experiential Learning: Dual Guidance for Utilization and Internalization 提出DGO双重引导优化框架，提升LLM在RLVR训练中的经验利用与内化能力 reinforcement learning large language model
14 ChargeFlow: Flow-Matching Refinement of Charge-Conditioned Electron Densities 提出ChargeFlow，通过流匹配细化电荷条件下的电子密度，加速材料科学计算。 flow matching

🔬 支柱九：具身大模型 (Embodied Foundation Models) (7 篇)

#题目一句话要点标签🔗⭐
15 Scaling Recurrence-aware Foundation Models for Clinical Records via Next-Visit Prediction RAVEN：基于复发感知的下一访问预测，扩展临床记录基础模型 foundation model
16 A Neuro-Symbolic System for Interpretable Multimodal Physiological Signals Integration in Human Fatigue Detection 提出一种神经符号系统，用于可解释的多模态生理信号融合，以检测人类疲劳。 multimodal
17 Adaptive decision-making for stochastic service network design 提出基于模拟退火和自适应代理模型的随机服务网络设计方法 multimodal
18 Forecasting with Guidance: Representation-Level Supervision for Time Series Forecasting ReGuider：利用表征级监督提升时间序列预测精度 foundation model
19 Lagrangian Relaxation Score-based Generation for Mixed Integer linear Programming 提出SRG框架以加速混合整数线性规划求解 zero-shot transfer
20 Understanding the Challenges in Iterative Generative Optimization with LLMs 揭示LLM迭代生成优化中的挑战，并提供实际应用指导 large language model
21 Diet Your LLM: Dimension-wise Global Pruning of LLMs via Merging Task-specific Importance Score DIET：通过融合任务特定重要性分数的维度级全局剪枝LLM方法 large language model

🔬 支柱一：机器人控制 (Robot Control) (1 篇)

#题目一句话要点标签🔗⭐
22 Towards Safe Learning-Based Non-Linear Model Predictive Control through Recurrent Neural Network Modeling 提出基于循环神经网络建模的安全学习非线性模型预测控制方法 MPC model predictive control

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

#题目一句话要点标签🔗⭐
23 Learning Response-Statistic Shifts and Parametric Roll Episodes from Wave--Vessel Time Series via LSTM Functional Models 提出基于LSTM函数模型的代理模型，用于学习波浪-船舶时序数据中的参数横摇事件和响应统计变化。 motion tracking

⬅️ 返回 cs.LG 首页 · 🏠 返回主页