List your company
ResearchFormerly WizWand

Hierarchical reasoning on symbolic sequences on Long ListOps (test)

62.75AccuracyLiquid-S4
13.942826.613939.28551.9561May 2, 2023
Updated 8mo ago

Evaluation results

MethodLinks
Liquid-S4Model Type=State Space Models2023.0562.75
MULTIRESNETModel Type=Convolution2023.0562.75
S5Model Type=State Space Models2023.0562.15
SGConv*Model Type=Convolution, Note=Built to mimic global convolution interpretation of S42023.0561.45
S4DModel Type=State Space Models2023.0560.52
S4Model Type=State Space Models2023.0559.6
DSSModel Type=State Space Models2023.0557.6
H-Transformer-1DModel Type=Attention2023.0549.53
CDILModel Type=Convolution2023.0544.05
ReformerModel Type=Attention2023.0537.27
Luna-256Model Type=Attention2023.0537.25
NyströmformerModel Type=Attention2023.0537.15
TransformerModel Type=Attention2023.0536.37
BigBirdModel Type=Attention2023.0536.05
LongformerModel Type=Attention2023.0535.63
FNetModel Type=Attention2023.0535.33
Sinkhorn TransformerModel Type=Attention2023.0533.67
PerformerModel Type=Attention2023.0518.01
Sparse TransformerModel Type=Attention2023.0517.07
Linear Trans.Model Type=Attention2023.0516.13
LinformerModel Type=Attention2023.0516.13
Local AttentionModel Type=Attention2023.0515.82