List your company
ResearchFormerly WizWand

Natural Language Inference on MultiNLI Matched

80.2AccuracyCAFE Ensemble
32.56844.93457.369.666Aug 7, 2017Jul 21, 2018Jul 4, 2019Jun 16, 2020May 30, 2021May 13, 2022Apr 27, 2023
Updated 8mo ago

Evaluation results

MethodLinks
CAFE Ensembleensemble_size=5 models2017.1280.2
Densely Interactive Inference Network2018.1179.2
CAFE2017.1278.7
Compare-Propagate Alignment-Factorized Encoders2018.1178.7
ESIM + Readsource=Weissenborn, 20172017.1277.8
ESIMsource=Weissenborn, 20172017.1276.3
Chen et al.2018.1174.9
Shortcut-Stacked Encoder2017.0874.6
Nie and Bansal2018.1174.6
Distance-Based Self-Attention Network2018.1174.1
aESIMimplementation=implemented on Keras2018.1273.9
ESIMimplementation=implemented on Keras2018.1273.4
ESIMsource=Williams et al., 2017 / Khot et al., 20182017.1272.4
Enhanced Sequential Inference Model2018.1172.4
biLSTM-Max Encoderreimplementation=true2017.0871.7
Stacked Bi-LSTMs + shortcut connections + max-poolinghidden_state_dim=300, bidirectional_dim=600, optimizer=Adam2018.1171.4
Directional Self-Attention Encoders2018.1171
DiSAN2017.0970.977
Bi-LSTM sentence encoder + max-pooling + attentionhidden_state_dim=300, bidirectional_dim=600, optimizer=Adam2018.1170.8
Bi-LSTM sentence encoder + max-poolinghidden_state_dim=300, bidirectional_dim=600, optimizer=Adam2018.1170.7
Stacked Bi-LSTMs + shortcut connections + max-pooling + attentionhidden_state_dim=300, bidirectional_dim=600, optimizer=Adam2018.1170.7
Bi-LSTM sentence encoder + attentionhidden_state_dim=300, bidirectional_dim=600, optimizer=Adam2018.1169.9
BiLSTM2017.1269.8
Bi-GRU sentence encoder + attentionhidden_state_dim=300, bidirectional_dim=600, optimizer=Adam2018.1169.4
Stacked Bi-GRUs + shortcut connections + max-pooling + attentionhidden_state_dim=300, bidirectional_dim=600, optimizer=Adam2018.1168.9
Stacked Bi-GRUs + shortcut connections + max-poolinghidden_state_dim=300, bidirectional_dim=600, optimizer=Adam2018.1168.5
SWEM-max2018.0568.2
SWEM-concat2018.0567.9
Bi-LSTM2017.0967.507
biLSTM Encoder2017.0867.5
BiLSTM2018.1167.5
Bi-GRU sentence encoder + max-pooling + attentionhidden_state_dim=300, bidirectional_dim=600, optimizer=Adam2018.1167.2
LSTMType=Bidirectional2018.0566.9
Bi-LSTMsource=baseline from [17]2018.1266.9
SWEM-aver2018.0566.5
Bi-GRU sentence encoder + max-poolinghidden_state_dim=300, bidirectional_dim=600, optimizer=Adam2018.1166.5
CBOW2017.0965.2
CBOW2017.0865.2
CBOW2017.1265.2
Continuous BOW (Averaging Word Embeddings)2018.1165.2
CNN2018.0565
CBOWsource=baseline from [17]2018.1264.8
LaMini-LLaMA# of params=7B2023.0463.8
LaMini-GPT-J# of params=6B2023.0457.7
Alpaca# of params=7B2023.0438.8
GPT-J# of params=6B2023.0437.4
Majority2017.1236.5
Most Frequent Class2018.1136.5
LLaMA# of params=7B2023.0434.4