cs.LG（2025-08-18）

📊 共 9 篇论文 | 🔗 5 篇有代码

🎯 兴趣领域导航

支柱九：具身大模型 (Embodied Foundation Models) (4 🔗2) 支柱二：RL算法与架构 (RL & Architecture) (3 🔗3) 支柱四：生成式动作 (Generative Motion) (1) 支柱五：交互与反应 (Interaction & Reaction) (1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (4 篇)

#题目一句话要点标签🔗⭐
1 Learning to Steer: Input-dependent Steering for Multimodal LLMs 提出L2S方法以解决多模态LLMs的输入依赖性引导问题 multimodal ✅
2 Z-Pruner: Post-Training Pruning of Large Language Models for Efficiency without Retraining 提出Z-Pruner以解决大语言模型后训练剪枝效率问题 large language model ✅
3 Generalization vs. Memorization in Autoregressive Deep Learning: Or, Examining Temporal Decay of Gradient Coherence 提出影响函数形式化以解决自回归深度学习中的泛化与记忆问题 foundation model
4 From Classical Probabilistic Latent Variable Models to Modern Generative AI: A Unified Perspective 提出统一视角将经典概率潜变量模型与现代生成AI相结合 large language model

🔬 支柱二：RL算法与架构 (RL & Architecture) (3 篇)

#题目一句话要点标签🔗⭐
5 Efficient Constraint-Aware Flow Matching via Randomized Exploration 提出高效的约束感知流匹配方法以解决样本生成问题 flow matching ✅
6 Flow Matching for Efficient and Scalable Data Assimilation 提出集成流过滤器以解决高维数据同化效率问题 flow matching ✅
7 MDPO: Overcoming the Training-Inference Divide of Masked Diffusion Language Models 提出MDPO以解决掩码扩散语言模型训练与推理不一致问题 reinforcement learning diffusion policy ✅

🔬 支柱四：生成式动作 (Generative Motion) (1 篇)

#题目一句话要点标签🔗⭐
8 Physically Plausible Data Augmentations for Wearable IMU-based Human Activity Recognition Using Physics Simulation 提出物理可行的数据增强方法以解决传感器基础的人体活动识别问题 physically plausible video-based pose estimation

🔬 支柱五：交互与反应 (Interaction & Reaction) (1 篇)

#题目一句话要点标签🔗⭐
9 Efficient and Verifiable Privacy-Preserving Convolutional Computation for CNN Inference with Untrusted Clouds 提出高效可验证的隐私保护卷积计算方案以解决云计算中的隐私泄露问题 OMOMO

⬅️ 返回 cs.LG 首页 · 🏠 返回主页