Towards Efficient Few-shot Graph Neural Architecture Search via Partitioning Gradient Contribution

Open in new window