Squirrel
Benchmarks
| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Node Classification | Squirrel | Accuracy83.49 | 500 | |
| Node classification | Squirrel (test) | Mean Accuracy76.01 | 234 | |
| Node Classification | Squirrel 60%/20%/20% random splits | Accuracy62.14 | 44 | |
| Node Classification | Squirrel (fixed) | Accuracy61.81 | 42 | |
| Node Classification | Squirrel (48/32/20) | Mean Accuracy74.1 | 40 | |
| Node Classification | Squirrel filtered | Accuracy41.8 | 35 | |
| Node Classification | Squirrel Standard (test) | Accuracy61.81 | 33 | |
| Node Classification | Squirrel (10 different splits) | Accuracy0.6181 | 27 | |
| Node Classification | Squirrel dense filtered | Accuracy42.66 | 26 | |
| Node Classification | Squirrel dense split original | Accuracy68.93 | 26 | |
| Node Classification | Squirrel Undirected (test) | Accuracy64.23 | 23 | |
| Node unlearning | Squirrel | Average Runtime (s)0.01 | 20 | |
| Node Regression | Squirrel | Normalized MAE Loss0.589 | 20 | |
| Node clustering | Squirrel (test) | Conductance0.24 | 15 | |
| Node Classification | Squirrel homophily: 0.22 (test) | Accuracy57.32 | 15 | |
| Node Classification | Squirrel Hnode = 0.22 | Accuracy61.48 | 14 | |
| Node Classification | Squirrel original (test) | Accuracy72.24 | 14 | |
| Node Classification | Squirrel Heterophilous (train val test) | Accuracy61.8 | 12 | |
| Node Classification | Squirrel heterophilic (test) | Accuracy68.76 | 12 | |
| Node Classification | Squirrel (full-supervised) | Mean Accuracy59.77 | 11 | |
| Node Classification | Squirrel semi-supervised (test) | Mean Accuracy30.91 | 11 | |
| Node-level Out-of-Distribution Detection | Squirrel (test) | AUROC (aleatoric)0.7502 | 10 | |
| Node Classification | Squirrel protocol (random) | Accuracy35.17 | 10 | |
| Node Classification | Squirrel (dense) | Mean Accuracy49.93 | 10 | |
| node-level task | Squirrel | Max GPU Memory (MB)1.436 | 9 |
Showing 25 of 41 rows