Natural Language Inference on MultiNLI mismatched (cross-domain) RepEval 2017 (test)
75.8AccuracyOur ESIM
Evaluation Results
| Method | Links | |
|---|---|---|
| Our ESIMimplementation=authors2017.08 | 75.8 | |
| Ensembledensemble=true, RepEval 2017 participant=true2017.08 | 74.9 | |
| Singleensemble=false, RepEval 2017 participant=true2017.08 | 73.6 | |
| Single (Input Gate)gate_type=input, RepEval 2017 participant=true2017.08 | 73.6 | |
| Single - Gated-Attgated_attention=removed2017.08 | 73.6 | |
| Shortcut-Stacked Encoder2017.08 | 73.6 | |
| YixinNie-UNC-NLPRepEval 2017 participant=true2017.08 | 73.5 | |
| Single - CharCNNcharacter_composition=removed2017.08 | 73.5 | |
| Single (Output Gate)gate_type=output2017.08 | 73.4 | |
| Single (Forget Gate)gate_type=forget2017.08 | 73.1 | |
| RivercornersRepEval 2017 participant=true, ensemble=true2017.08 | 72.8 | |
| ESIM2017.08 | 72.1 | |
| RivercornersRepEval 2017 participant=true, ensemble=false2017.08 | 72.1 | |
| DiSAN2017.09 | 71.402 | |
| biLSTM-Max Encoderreimplementation=true2017.08 | 71.2 | |
| LCT-MALTARepEval 2017 participant=true2017.08 | 70.8 | |
| Single - AbsDiff/Productelementwise_vector_comparison=removed2017.08 | 69.2 | |
| TALP-UPCRepEval 2017 participant=true2017.08 | 68.2 | |
| Bi-LSTM2017.09 | 67.248 | |
| biLSTM Encoder2017.08 | 67.1 | |
| BiLSTM2017.08 | 66.9 | |
| Single - Word Embeddingword_level_embedding=removed2017.08 | 66 | |
| CBOW2017.09 | 64.759 | |
| CBOW2017.08 | 64.6 | |
| CBOW2017.08 | 64.5 |