List your company
ResearchFormerly WizWand

Machine Reading Comprehension on Daily Mail (test)

80.9AccuracyGA Reader
33.68445.94258.270.458Mar 4, 2016Apr 14, 2016May 25, 2016Jul 5, 2016Aug 15, 2016Sep 25, 2016Nov 5, 2016
Updated 8mo ago

Evaluation results

MethodLinks
GA Readervariant=GA (update L(w))2016.0680.9
GA Readervariant=GA (+feature, update L(w))2016.0680
BiDAFensemble=false2016.1179.6
BIDAF2016.0679.6
GA Readervariant=GA (fix L(w))2016.0679.6
GA Readervariant=GA (+feature, fix L(w))2016.0679.3
Stanford ARensemble=true2016.1179.2
Neural netRelabeling=true, Ensemble=true2016.0679.2
Stanford AR (relabeling)ensemble=true2016.0679.2
Neural netEnsemble=true2016.0678.7
GAReaderensemble=true2016.1178.1
AS ReaderAggregation=greedy ensemble2016.0377.7
AS Readerensemble=true2016.1177.7
AS Readerensemble=true2016.0677.7
AS ReaderAggregation=avg ensemble2016.0377.1
Stanford ARensemble=false2016.1176.6
ReasoNetensemble=false2016.1176.6
Neural netRelabeling=true2016.0676.6
Stanford AR (relabeling)2016.0676.6
ReasoNet2016.0676.6
ReasoNetTmax=52016.0976.6
Neural net2016.0676
Chen et al. (2016)Model configuration=single model2016.0375.8
Stanford AR2016.0975.8
GAReaderensemble=false2016.1175.7
GA Readervariant=GA--2016.0675.7
GA Reader2016.0975.7
AS ReaderModel configuration=single model2016.0373.9
AS Readerensemble=false2016.1173.9
AS Reader2016.0673.9
AS Reader2016.0973.9
AS ReaderAggregation=avg for top 20%2016.0373.5
Attentive ReaderModel type=single model2016.0369
Attentive Readerensemble=false2016.1169
Attentive Reader2016.0669
Attentive Reader2016.0669
Attentive Reader2016.0969
Entity-centric Classifier2016.0668.3
Impatient ReaderModel type=single model2016.0368
Impatient Reader2016.0668
Impatient Reader2016.0668
Deep LSTM Reader2016.0662.2
Deep LSTM Reader2016.0662.2
Deep LSTM Reader2016.0962.2
Word distance model2016.0655.5
Frame-semantic model2016.0635.5