Learning General Purpose Distributed Sentence Representations via Large Scale Multi-task Learning

ICLR 2018 Sandeep SubramanianAdam TrischlerYoshua BengioChristopher J Pal

A lot of the recent success in natural language processing (NLP) has been driven by distributed vector representations of words trained on large amounts of text in an unsupervised manner. These representations are typically used as general purpose features for words across a range of NLP problems... (read more)

PDF Abstract

Evaluation results from the paper


Task Dataset Model Metric name Metric value Global rank Compare
Natural Language Inference MultiNLI GenSen Matched 71.4 # 11
Natural Language Inference MultiNLI GenSen Mismatched 71.3 # 10
Paraphrase Identification Quora Question Pairs GenSen Accuracy 87.01 # 7
Semantic Textual Similarity SentEval GenSen MRPC 78.6/84.4 # 1
Semantic Textual Similarity SentEval GenSen SICK-R 0.888 # 1
Semantic Textual Similarity SentEval GenSen SICK-E 87.8 # 1
Semantic Textual Similarity SentEval GenSen STS 78.9/78.6 # 1