Hierarchical Reasoning on ListOps Long Range Arena (test)
63.04AccuracyMS-SSM (S6)
Evaluation Results
| Method | Links | |
|---|---|---|
| MS-SSM (S6)Architecture type=Convolutions + SSMs2025.12 | 63.04 | |
| MS-SSM (S4)Architecture type=Convolutions + SSMs2025.12 | 62.83 | |
| MULTIRESNETArchitecture type=Convolutions2025.12 | 62.75 | |
| Liquid-S4Architecture type=SSMs2025.12 | 62.75 | |
| S5Architecture type=SSMs2025.12 | 62.15 | |
| SGConvArchitecture type=Convolutions2025.12 | 61.45 | |
| S4DArchitecture type=SSMs2025.12 | 60.52 | |
| S4Architecture type=SSMs2025.12 | 59.6 | |
| DSSArchitecture type=SSMs2025.12 | 57.6 | |
| MambaArchitecture type=SSMs, Model size=double model size (2x Param)2025.12 | 49.63 | |
| H-Transformer-1DArchitecture type=Transformers2025.12 | 49.53 | |
| CDILArchitecture type=Convolutions2025.12 | 44.05 | |
| MambaArchitecture type=SSMs, State size=double state size (2x State)2025.12 | 42.14 | |
| MambaArchitecture type=SSMs2025.12 | 38.02 | |
| ReformerArchitecture type=Transformers2025.12 | 37.27 | |
| Luna-256Architecture type=Transformers2025.12 | 37.25 | |
| TransformerArchitecture type=Transformers2025.12 | 36.37 | |
| BigBirdArchitecture type=Transformers2025.12 | 36.05 | |
| LongformerArchitecture type=Transformers2025.12 | 35.63 | |
| Sinkhorn TransformerArchitecture type=Transformers2025.12 | 33.67 | |
| GriffinArchitecture type=SSMs2025.12 | 32.34 | |
| PerformerArchitecture type=Transformers2025.12 | 18.01 | |
| Sparse TransformerArchitecture type=Transformers2025.12 | 17.07 | |
| Linear Trans.Architecture type=Transformers2025.12 | 16.13 | |
| LinformerArchitecture type=Transformers2025.12 | 16.13 | |
| Local AttentionArchitecture type=Transformers2025.12 | 15.82 |