Our new X account is live! Follow @wizwand_team for updates

Natural Language Inference on MultiNLI matched (test)

85.38Accuracy

BERT-base + PoE

Updated 4d ago

Evaluation Results

Method	Links
BERT-base + PoE 2022.03		85.38
BERT-base 2022.03		85.12
BERT-base + PoE 2022.03		84.69
PoE 2022.03		84.11
BERT-base 2022.03		84.11
Regularized-conf 2022.03		84.1
BERT-base 2022.03		83.41
BERT-base 2022.03		83.27
DRCN+ELMo 2018.05		82.3
LM-Transformer 2018.05		82.1
DRCN 2018.05		80.6
CAFE 2018.05		80.2
DIIN 2017.09		80
DIIN 2018.05		80
Learned-Mixin 2022.03		79.5
DRCN 2018.05		79.1
DIIN 2017.09		78.8
DIIN 2018.05		78.8
CAFE 2018.05		78.7
CAFE 2017.11		78.7
DIIN 2017.11		77.8
KIM 2017.11		77.2
ESIM 2017.11		76.8
Gated-Att BiLSTM 2017.09		74.9
Shortcut-Stacked encoder 2017.09		74.6
Shortcut stacked BiLSTM 2018.09		74.6
2-layer Bi-CAS-LSTM 2018.09		74.6
SS BILSTM 2017.11		74.6
ESIM GloVe 2017.06		74.58
Shortcut-Stacked BiLSTM encoders 2017.12		74.5
Shortcut-Stacked BiLSTM 2018.08		74.5
3-layer Bi-CAS-LSTM 2018.09		74.2
Distance-based Self-Attention Network 2017.12		74.1
2-layer CAS-LSTM 2018.09		74
BiLSTM with generalized pooling 2018.09		73.8
3-layer CAS-LSTM 2018.09		73.8
HBMP 2018.08		73.7
Deep Gated Attn. BiLSTM encoders 2017.12		73.5
Deep Gated Attn. BiLSTM encoders 2018.08		73.5
BiLSTM with gated pooling 2018.09		73.5
Gated BiLSTM 2017.11		73.5
Gated-Att BiLSTM 2017.09		73.2
ESIM 2017.09		72.3
ESIM 2018.05		72.3
InnerAtt 2017.09		72.2
BiLSTM + Inner-attention 2017.12		72.1
InnerAtt 2017.09		72.1
BiLSTM + Inner-attention 2018.08		72.1
ESIM dictionary 2017.06		71.45
Directional Self-Attention Network 2017.12		71
DiSAN 2017.11		71
BiLSTM + enhanced embedding + max pooling 2017.12		70.7
BiLSTM + enh embed + max pooling 2018.08		70.7
ESIM spelling 2017.06		68.89
ESIM baseline 2017.06		68.55
Cha-level Intra-attention BiLSTM encoders 2017.12		67.9
BiLSTM 2017.12		67.5
BiLSTM 2018.08		67.5
BiLSTM 2017.09		67
BiLSTM 2018.09		66.9
BILSTM 2017.11		66.9
CBOW 2017.12		66.2
CBOW 2018.08		66.2
CBOW 2018.09		64.8
CBOW 2017.11		64.8