| 1 |
A million-scale dataset and generalizable foundation model for nanomaterial-protein interactions |
提出NanoPro-3M数据集与NanoProFormer模型，用于预测纳米材料-蛋白质相互作用。 |
representation learning foundation model multimodal |
|
|
| 2 |
LLaPipe: LLM-Guided Reinforcement Learning for Automated Data Preparation Pipeline Construction |
LLaPipe：利用LLM指导的强化学习构建自动化数据准备流水线 |
reinforcement learning distillation large language model |
|
|
| 3 |
Reframing attention as a reinforcement learning problem for causal discovery |
提出Causal Process Model，将注意力机制重构为强化学习问题以进行因果发现。 |
reinforcement learning deep reinforcement learning representation learning |
|
|
| 4 |
SoftPipe: A Soft-Guided Reinforcement Learning Framework for Automated Data Preparation |
提出SoftPipe框架以解决数据准备中的搜索空间问题 |
reinforcement learning large language model |
|
|
| 5 |
State Space Models Naturally Produce Traveling Waves, Time Cells, and Scale to Abstract Cognitive Functions |
提出基于状态空间模型(SSM)的框架，统一神经元动力学与认知功能，解释时间细胞涌现。 |
reinforcement learning SSM state space model |
|
|
| 6 |
Preference-based Multi-Objective Reinforcement Learning |
提出基于偏好的多目标强化学习，解决复杂任务中奖励函数难以设计的问题 |
reinforcement learning reward design |
|
|
| 7 |
Toward Temporal Causal Representation Learning with Tensor Decomposition |
提出CaRTeD框架，结合张量分解与时序因果表示学习，处理高维不等长时序数据。 |
representation learning |
|
|
| 8 |
BikeVAE-GNN: A Variational Autoencoder-Augmented Hybrid Graph Neural Network for Sparse Bicycle Volume Estimation |
提出BikeVAE-GNN，解决城市自行车网络中稀疏流量估计问题 |
MAE spatial relationship |
|
|
| 9 |
Dual-Center Graph Clustering with Neighbor Distribution |
提出基于邻居分布的双中心图聚类方法，提升图聚类性能。 |
representation learning contrastive learning |
|
|