It's my honor to read your excellent paper which has broadened my view greatly. You conduct associated experiments on two OGB node property prediction datasets, namely, ogbn-arxiv and ogbn-products. I think the two datasets are somewhat small-scale or medium-scale, did you try to conduct experiments on ogbn-papers100M dataset?
It's my honor to read your excellent paper which has broadened my view greatly. You conduct associated experiments on two OGB node property prediction datasets, namely, ogbn-arxiv and ogbn-products. I think the two datasets are somewhat small-scale or medium-scale, did you try to conduct experiments on ogbn-papers100M dataset?