. 2025 Feb 25;11:e2706. doi: 10.7717/peerj-cs.2706

Table 5. The impact of varying MLP depths on GNN-A²’s performance.

To enhance the clarity of the experimental results, the optimal results for each dataset are highlighted in bold, and the suboptimal ones are indicated with underlining.

Model	MovieLens 1M		Book-crossing		Taobao
	AUC	NDCG@10	AUC	NDCG@10	AUC	NDCG@10
GNN-A²-0	0.9050	0.9421	0.8296	0.9015	0.6702	0.1467
GNN-A²-1	0.9101	0.9506	0.8400	0.9137	0.6715	0.1526
GNN-A²-2	0.9091	0.9489	0.8387	0.9090	0.6709	0.1503
GNN-A²-3	0.9076	0.9456	0.8367	0.9073	0.6704	0.1489

Table 5. The impact of varying MLP depths on GNN-A2’s performance.

Table 5. The impact of varying MLP depths on GNN-A²’s performance.