List your company
ResearchFormerly WizWand

Tense Reinflection on tense reinflection without auxiliary do (test)

100Full AccuracyTf
-2.75223.92450.677.276Nov 5, 2025
Updated 4mo ago

Evaluation results

MethodLinks
TfSource=Ahuja et al. (2025), Training objective=language modeling objective2025.11100
TfSource=Qin et al. (2025)2025.11100
TfArchitecture=Transformer, Ground-truth parse trees=true2025.1199.9
Stack Attention TransformerArchitecture=Tf+Nd2025.1199.8
Tf+Nd+NdArchitecture=Transformer with Double Stack2025.1199.7
Tf+SupArchitecture=Transformer with Supervision2025.1199.3
Tf+Sup+SupArchitecture=Transformer with Double Supervision2025.1198.4
LSTMSource=McCoy et al. (2020)2025.1196
Tree-GRUSource=McCoy et al. (2020)2025.1196
ON-LSTMSource=McCoy et al. (2020)2025.1195
Stack RNNArchitecture=RNN+Nd2025.115.9
RNNArchitecture=RNN2025.115.1
LSTMArchitecture=LSTM2025.112.9
RNN+SupArchitecture=RNN with Supervision2025.112.4
RNN+Nd+RArchitecture=Stack RNN with Reinforcement2025.112.4
LSTM+Sup+RArchitecture=LSTM with Supervision and Reinforcement2025.111.8
RNN+Sup+RArchitecture=RNN with Supervision and Reinforcement2025.111.7
LSTM+Nd+RArchitecture=Stack LSTM with Reinforcement2025.111.7
LSTM+SupArchitecture=LSTM with Supervision2025.111.6
Stack LSTMArchitecture=LSTM+Nd2025.111.2