cs.LG（2026-06-16）

📊 共 31 篇论文 | 🔗 2 篇有代码

🎯 兴趣领域导航

支柱二：RL算法与架构 (RL & Architecture) (18 🔗2) 支柱九：具身大模型 (Embodied Foundation Models) (8) 支柱一：机器人控制 (Robot Control) (2) 支柱八：物理动画 (Physics-based Animation) (2) 支柱四：生成式动作 (Generative Motion) (1)

🔬 支柱二：RL算法与架构 (RL & Architecture) (18 篇)

#题目一句话要点标签🔗⭐
1 Ternary Mamba: Grouped Quantization-Aware Training of W1.58A16 State Space Models 提出Ternary Mamba以解决SSM边缘部署的内存限制问题 Mamba SSM state space model
2 Memory-Efficient Meta-Reinforcement Learning for Adaptive Safety-Critical Control in Adversarial Spacecraft Proximity Operations 提出记忆高效的元强化学习以解决自适应安全关键控制问题 reinforcement learning PPO SAC
3 Looped World Models 提出LoopWM以解决长时间模拟中的计算效率问题 world model world models
4 Deep Reinforcement Learning for Minimum Zero-Forcing Sets 提出SD-ZFS框架以解决最小零强迫集问题 reinforcement learning deep reinforcement learning
5 EnvRL: Learn from Environment Dynamics in Agentic Reinforcement Learning 提出EnvRL框架以改善长时间任务中的稀疏奖励问题 reinforcement learning policy learning large language model
6 Continuous-time Optimal Stopping through Deep Reinforcement Learning 提出CARLOS算法以解决连续时间最优停止问题 reinforcement learning deep reinforcement learning
7 OmniPlan: An Adaptive Framework for Timely and Near-Optimal Network Planning Optimization 提出OmniPlan框架以解决网络规划优化问题 reinforcement learning deep reinforcement learning DRL
8 S4oP: Operator-level Pruning of Structured State Space Models for Resource-Constrained Devices 提出S4oP以解决资源受限设备上的SSM模型效率问题 SSM state space model
9 Reversal Q-Learning 提出Reversal Q-Learning以提升离线强化学习性能 reinforcement learning offline RL offline reinforcement learning
10 Multi-Adapter PPO: A Cross-Attention Enhanced Wavelength Selection Framework for LIBS Quantitative Analysis 提出Multi-Adapter PPO框架以解决LIBS定量分析中的波长选择问题 reinforcement learning PPO ✅
11 Rethinking Dataset Distillation for Classification: Do Distilled Sets Outperform Coresets? 评估数据集蒸馏在分类任务中的有效性与核心集的比较 distillation
12 Learning Fair Pareto-Optimal Policies in Multi-Objective Reinforcement Learning 提出公平的Pareto最优策略以解决多目标强化学习中的公平性问题 reinforcement learning
13 From Reasoning Traces to Reusable Modules: Understanding Compositional Generalization in Language Model Reasoning 提出层次潜在选择模型以增强语言模型的组合泛化能力 reinforcement learning large language model
14 Continual Self-Improvement with Lightweight Experiential Latent Memories 提出轻量级经验潜在记忆以实现持续自我改进 reinforcement learning large language model
15 Confusion-Aware Transfer Teacher Curriculum Learning Framework: Disentangling Scoring and Pacing Effects 提出混淆感知转移教师课程学习框架以解决评分与节奏效应问题 curriculum learning
16 Expanding SPHERE-JEPA: A Family of Statistical Regularizers for the Hypersphere 提出全维统计正则化方法以解决自监督学习中的表示崩溃问题 JEPA
17 MGUP: A Momentum-Gradient Alignment Update Policy for Stochastic Optimization 提出MGUP以解决随机优化中的选择性更新问题 MAE large language model ✅
18 ResAware: Cross-Environment Website Fingerprinting via Resource-Privileged Distillation 提出ResAware以解决跨环境网站指纹识别的鲁棒性问题 distillation

🔬 支柱九：具身大模型 (Embodied Foundation Models) (8 篇)

#题目一句话要点标签🔗⭐
19 Half a Link can Be Enough to Predict a Whole Link: Understanding Generalization in Knowledge Graph Foundation Models 提出半链接概念以提升知识图谱基础模型的泛化能力 foundation model
20 Learning to Refine Hidden States for Reliable LLM Reasoning 提出ReLAR框架以解决大型语言模型推理不稳定问题 large language model chain-of-thought
21 Evaluating Open-Source LLMs for Multi-Label ATT&CK Technique Classification on CTI Reports 构建数据集评估开源LLM在CTI报告中的ATT&CK分类能力 large language model
22 AnchorKV: Safety-Aware KV Cache Compression via Soft Penalty with a Refusal Anchor 提出AnchorKV以解决KV缓存压缩中的安全性问题 large language model
23 Handling Feature Heterogeneity with Learnable Graph Patches 提出可学习图块以解决图数据特征异质性问题 foundation model
24 TuneAhead: Predicting Fine-tuning Performance Before Full Training Begins 提出TUNEAHEAD以预测微调性能，降低训练成本 large language model
25 Public transit gains and spatially uneven travel demand changes after NYC congestion pricing 提出基于时间序列模型的需求预测框架以评估纽约市拥堵定价政策影响 foundation model
26 Online LLM Selection via Constrained Bandits with Time-Varying Demand 提出基于约束博弈的在线LLM选择方法以应对动态需求问题 large language model

🔬 支柱一：机器人控制 (Robot Control) (2 篇)

#题目一句话要点标签🔗⭐
27 Kolmogorov Regression for Robust Diffusion Policies 提出Kolmogorov回归以解决扩散策略的鲁棒性问题 manipulation diffusion policy
28 Dimensionality Controls When Modularity Helps in Continual Learning 提出模块化架构以提升持续学习中的表现与稳定性 manipulation

🔬 支柱八：物理动画 (Physics-based Animation) (2 篇)

#题目一句话要点标签🔗⭐
29 ASTEROID: A Spatiotemporal Information Transformer for Forecasting Multi-Step Time Series of Molecular Dynamics 提出ASTEROID以解决分子动力学多步预测问题 spatiotemporal
30 SpatioTemporal Causal Network Diagnostics for Geographic Tipping Point Early Warning 提出时空因果网络诊断以解决地理临界点预警问题 spatiotemporal

🔬 支柱四：生成式动作 (Generative Motion) (1 篇)

#题目一句话要点标签🔗⭐
31 Recursive Scaling in Masked Diffusion Models 提出递归掩蔽扩散模型以提升参数效率 MDM

⬅️ 返回 cs.LG 首页 · 🏠 返回主页