reinforce

3 papers

• REINFORCE Converges to Optimal Policies with Any Learning Rate
• REOrdering Patches Improves Vision Models
• UniteFormer: Unifying Node and Edge Modalities in Transformers for Vehicle Routing Problems