cs.LG（2025-11-14）

📊 共 24 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) (11 🔗2) 支柱九：具身大模型 (Embodied Foundation Models) (8 🔗1) 支柱一：机器人控制 (Robot Control) (4) 支柱八：物理动画 (Physics-based Animation) (1)

🔬 支柱二：RL算法与架构 (RL & Architecture) (11 篇)

#题目一句话要点标签🔗⭐
1 Enhancing Robustness of Offline Reinforcement Learning Under Data Corruption via Sharpness-Aware Minimization 提出基于锐度感知最小化的离线强化学习方法，提升数据损坏下的鲁棒性 reinforcement learning offline RL offline reinforcement learning
2 PROF: An LLM-based Reward Code Preference Optimization Framework for Offline Imitation Learning PROF：基于LLM的离线模仿学习奖励代码偏好优化框架 policy learning imitation learning large language model
3 Toward Scalable Early Cancer Detection: Evaluating EHR-Based Predictive Models Against Traditional Screening Criteria 利用电子病历预测模型实现可扩展的早期癌症检测，优于传统筛查标准。 predictive model foundation model
4 LoRaCompass: Robust Reinforcement Learning to Efficiently Search for a LoRa Tag LoRaCompass：基于鲁棒强化学习的高效LoRa标签搜索方法 reinforcement learning distillation
5 Low-Bit, High-Fidelity: Optimal Transport Quantization for Flow Matching 提出基于最优传输量化的Flow Matching模型压缩方法，实现低比特高保真生成。 flow matching
6 Better LLM Reasoning via Dual-Play PasoDoble：一种基于双人对抗博弈的无监督LLM推理能力提升方法 reinforcement learning large language model ✅
7 Dynamic Temperature Scheduler for Knowledge Distillation 提出动态温度调度器DTS，通过教师-学生模型差异自适应调整知识蒸馏温度。 distillation ✅
8 Credal Ensemble Distillation for Uncertainty Quantification 提出Credal Ensemble Distillation (CED)框架，用于深度集成模型的知识蒸馏和不确定性量化。 distillation
9 HealSplit: Towards Self-Healing through Adversarial Distillation in Split Federated Learning HealSplit：面向分割联邦学习，通过对抗蒸馏实现自愈的数据中毒防御。 distillation
10 Efficient Reinforcement Learning for Zero-Shot Coordination in Evolving Games 提出ScaPT框架，解决演化博弈中零样本协同的计算资源瓶颈问题 reinforcement learning
11 Flow matching-based generative models for MIMO channel estimation 提出基于Flow Matching的MIMO信道估计算法，加速信道状态信息获取。 flow matching

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

#题目一句话要点标签🔗⭐
12 A Systematic Study of Model Extraction Attacks on Graph Foundation Models 针对图基础模型的模型提取攻击系统性研究，揭示其安全风险。 foundation model multimodal
13 Adaptive Redundancy Regulation for Balanced Multimodal Information Refinement 提出RedReg，通过自适应冗余调节实现平衡的多模态信息精炼。 multimodal ✅
14 Architecting software monitors for control-flow anomaly detection through large language models and conformance checking 提出基于大语言模型和一致性检验的软件监控架构，用于控制流异常检测 large language model
15 Leveraging Exogenous Signals for Hydrology Time Series Forecasting 利用外生信号改进水文时间序列预测，优于现有基础模型 foundation model
16 When Genes Speak: A Semantic-Guided Framework for Spatially Resolved Transcriptomics Data Clustering SemST：提出一种语义引导的深度学习框架，用于空间转录组数据聚类。 large language model
17 Fast and Expressive Multi-Token Prediction with Probabilistic Circuits 提出基于概率电路的MTPC框架，加速字节级LLM生成并保持性能。 large language model
18 SMART: A Surrogate Model for Predicting Application Runtime in Dragonfly Systems 提出SMART模型以预测Dragonfly系统中的应用运行时间 large language model
19 CAT-Net: A Cross-Attention Tone Network for Cross-Subject EEG-EMG Fusion Tone Decoding CAT-Net：一种用于跨个体脑电-肌电融合声调解码的跨注意力网络 multimodal

🔬 支柱一：机器人控制 (Robot Control) (4 篇)

#题目一句话要点标签🔗⭐
20 Multi-Phase Spacecraft Trajectory Optimization via Transformer-Based Reinforcement Learning 提出基于Transformer的强化学习框架，解决航天器多阶段轨迹优化问题。 trajectory optimization reinforcement learning PPO
21 Robustness of LLM-enabled vehicle trajectory prediction under data security threats 针对LLM车辆轨迹预测，提出单特征差分进化攻击以评估其数据安全鲁棒性 manipulation physically plausible large language model
22 Adaptive Intrusion Detection for Evolving RPL IoT Attacks Using Incremental Learning 提出基于增量学习的自适应入侵检测方法，应对演进的RPL IoT攻击。 manipulation
23 Differentiation Strategies for Acoustic Inverse Problems: Admittance Estimation and Shape Optimization 提出基于可微编程的声学反问题求解策略，用于导纳估计和形状优化。 manipulation

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

#题目一句话要点标签🔗⭐
24 MoCap2Radar: A Spatiotemporal Transformer for Synthesizing Micro-Doppler Radar Signatures from Motion Capture MoCap2Radar：利用时空Transformer从动作捕捉数据合成微多普勒雷达信号 spatiotemporal

⬅️ 返回 cs.LG 首页 · 🏠 返回主页