cs.LG（2026-08-13）

📊 共 28 篇论文 | 🔗 3 篇有代码

🎯 兴趣领域导航

支柱二：RL算法与架构 (RL & Architecture) (16 🔗1) 支柱九：具身大模型 (Embodied Foundation Models) (10 🔗2) 支柱七：动作重定向 (Motion Retargeting) (1) 支柱三：空间感知与语义 (Perception & Semantics) (1)

🔬 支柱二：RL算法与架构 (RL & Architecture) (16 篇)

#题目一句话要点标签🔗⭐
1 Diagnosing JEPA World Models with Action-Conditioned Predictive Consistency 提出动作条件预测一致性以诊断JEPA世界模型的视觉扰动问题 world model world models JEPA
2 Intern-S2-Preview: Scientific Agentic Foundation Model 提出Intern-S2-Preview以解决科学发现中的多模态理解问题 reinforcement learning distillation foundation model
3 Towards Socially Compliant Navigation in Deep Reinforcement Learning via Proxemics-Based Reward Modeling 提出基于亲密距离的奖励模型以解决社交合规导航问题 reinforcement learning deep reinforcement learning DRL
4 Intervention-Aware Clinical World Model for Post-Op Outcome Forecasting in Cardiology 提出干预感知临床世界模型以预测心脏病术后结果 world model world models MAE
5 The Objective Is the Bottleneck: Latent World Models Encode What Their Planners Cannot Use 提出新的目标函数以解决长时间规划中的瓶颈问题 world model worldmodel world models
6 CardioState-JEPA: Delay-Aware Cross-Modal Learning of a Shared Cardiac Representation 提出CardioState-JEPA以解决心脏信号多模态学习问题 JEPA Joint-Embedding Predictive Architecture joint-embedding predictive architecture
7 The Impact of Temporal Context Length and Encoding Strategies on Self-Supervised ECG Representation Learning 提出基于时间上下文长度与编码策略的自监督ECG表示学习方法 representation learning foundation model ✅
8 Beyond Outcome Rewards: Step-Level Self-Distilled Policy Optimization for Deep Search Agents 提出步骤级自蒸馏策略优化以解决深度搜索代理的稀疏奖励问题 reinforcement learning teacher-student distillation
9 FlowLOB: Efficient and Controllable Limit Order Book Generation with Flow Matching 提出FlowLOB以解决限价单簿生成效率与可控性问题 flow matching
10 Latent On-Policy Self-Distillation 提出潜在的在线自蒸馏方法以解决自我进化AI的学习效率问题 distillation
11 Online Inference for Quantile Temporal Difference Learning in Distributional Reinforcement Learning 提出在线推断方法以优化分布式强化学习中的量化时间差学习 reinforcement learning
12 I-SDPO: Instance-Level Adaptive Self-Distillation Policy Optimization 提出I-SDPO以解决自蒸馏策略优化中的偏差问题 distillation
13 Dynamic Multi-Depot Vehicle Routing with Online Requests: Event-Driven Transformer--DRL and Rolling-Horizon Benchmarking 提出事件驱动的动态多仓库车辆调度框架以应对在线请求问题 DRL PPO behavior cloning
14 The Query Knows What to Forget: A Second Erase Direction for Linear Attention 提出查询导向的擦除方向以解决线性注意力中的干扰问题 linear attention
15 Contrastive Learning for Interpretable Anomaly Detection at Collider Experiments 提出ORCA以解决对撞机实验中的可解释异常检测问题 contrastive learning
16 Online Inference for Quantile Temporal Difference Learning in Distributional Reinforcement Learning 提出在线推断方法以优化分布式强化学习中的量化时间差学习 reinforcement learning

🔬 支柱九：具身大模型 (Embodied Foundation Models) (10 篇)

#题目一句话要点标签🔗⭐
17 Foundation models for movement data: Are they ready for prime-time? 评估基础模型在运动数据中的应用以提升健康监测效果 foundation model
18 Into the ORBIT for Time Series: Training Regimes for Foundation Models 提出ORBIT训练范式以优化时间序列基础模型的训练 foundation model
19 HiRoute: Hierarchical Routed Prompt Tuning for Safety Alignment of Large Language Models 提出HiRoute以解决大型语言模型安全对齐问题 large language model
20 A Cloud-Edge System for Multimodal Clinical Screening in Resource-Constrained Rural Settings 提出云边协同系统以解决资源匮乏农村地区的多模态临床筛查问题 multimodal
21 Reduced Matrix Multiplication: Input-Adaptive Matrix-Product Reduction for LLM Inference 提出RMM以解决Transformer推理成本高的问题 multimodal
22 Balanced Adaptive Prototype Selection for Scalable TabPFN Inference on Large-Scale Tabular Data 提出平衡自适应原型选择以解决大规模表格数据推理问题 foundation model
23 MAG: MAnifold Guided Semi-Supervised Multi-modal In-Context Learning 提出MAG框架以提升多模态少样本学习效果 large language model
24 EEG-PRISM: Physiologically-Grounded Interpretability of Predictions by EEG Foundation Models 提出EEG-PRISM以解决EEG基础模型可解释性不足问题 foundation model
25 On the Brittleness of Maximum Likelihood Estimation for Gaussian Process Hyperparameter Optimization 评估最大似然估计在高斯过程超参数优化中的脆弱性 foundation model ✅
26 Language-Specific Gaps in AI Safety Training Datasets 提出多语言安全数据集审计方法以解决语言特定差距问题 large language model ✅

🔬 支柱七：动作重定向 (Motion Retargeting) (1 篇)

#题目一句话要点标签🔗⭐
27 Beyond Simulated Benchmarks: Evaluating Motion Representations for Fall Detection Under Real-World Data Scarcity 提出运动表示方法以解决真实世界跌倒检测数据稀缺问题 motion representation foundation model

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

#题目一句话要点标签🔗⭐
28 Architecture and Affordances of PLAUD: Performative Latents and Unsupervised DDSP 提出PLAUD以增强现场电子音乐表演的合成能力 affordance

⬅️ 返回 cs.LG 首页 · 🏠 返回主页