cs.LG（2026-02-18）

📊 共 16 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) (10) 支柱九：具身大模型 (Embodied Foundation Models) (5 🔗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) (10 篇)

#题目一句话要点标签🔗⭐
1 HiPER: Hierarchical Reinforcement Learning with Explicit Credit Assignment for Large Language Model Agents HiPER：通过显式信用分配的分层强化学习提升LLM Agent性能 reinforcement learning large language model
2 Capacity-constrained demand response in smart grids using deep reinforcement learning 提出基于深度强化学习的容量约束需求响应方法，优化智能电网资源分配 reinforcement learning deep reinforcement learning
3 Reinforcement Learning for Parameterized Quantum State Preparation: A Comparative Study 提出基于强化学习的参数化量子态制备方法以提升量子电路合成效率 reinforcement learning PPO
4 Factored Latent Action World Models 提出分解潜在动作模型以解决复杂环境中的控制问题 policy learning world model
5 Causality is Key for Interpretability Claims to Generalise 利用因果关系提升LLM可解释性研究的泛化能力 representation learning large language model
6 Vulnerability Analysis of Safe Reinforcement Learning via Inverse Constrained Reinforcement Learning 提出基于逆约束强化学习的安全强化学习策略脆弱性分析框架 reinforcement learning
7 Intra-Fairness Dynamics: The Bias Spillover Effect in Targeted LLM Alignment 研究揭示LLM性别对齐中的偏见溢出效应，强调多属性公平性评估的重要性 direct preference optimization large language model
8 Geometric Neural Operators via Lie Group-Constrained Latent Dynamics 提出基于李群约束流形神经网络算子，提升偏微分方程长期预测稳定性 latent dynamics
9 Graphon Mean-Field Subsampling for Cooperative Heterogeneous Multi-Agent Reinforcement Learning 提出GMFS框架以解决异质多智能体强化学习中的协调问题 reinforcement learning
10 A Scalable Approach to Solving Simulation-Based Network Security Games 提出MetaDOAR，通过可扩展的分层策略学习解决大规模网络安全博弈问题。 reinforcement learning policy learning

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

#题目一句话要点标签🔗⭐
11 Parameter-free representations outperform single-cell foundation models on downstream benchmarks 提出无参数表示方法以超越单细胞基础模型的下游任务表现 foundation model
12 Retrieval-Augmented Foundation Models for Matched Molecular Pair Transformations to Recapitulate Medicinal Chemistry Intuition 提出MMPT-RAG，利用检索增强的分子生成模型，模拟药物化学家的直觉进行分子改造。 foundation model
13 A Systematic Evaluation of Sample-Level Tokenization Strategies for MEG Foundation Models 系统评估MEG脑电数据样本级Tokenization策略对神经科学大模型的影响 foundation model ✅
14 ModalImmune: Immunity Driven Unlearning via Self Destructive Training ModalImmune：提出一种自毁训练框架，增强多模态系统对模态缺失的鲁棒性 multimodal
15 ASPEN: Spectral-Temporal Fusion for Cross-Subject Brain Decoding 提出ASPEN：通过频谱-时间融合实现跨被试脑电解码 multimodal

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

#题目一句话要点标签🔗⭐
16 Multi-Class Boundary Extraction from Implicit Representations 提出一种多类别隐式表示的二维边界提取算法，保证拓扑正确性和水密性。 implicit representation

⬅️ 返回 cs.LG 首页 · 🏠 返回主页