A controlled comparison shows translation and image-grounding objectives produce better sentence representations than language modeling on moderate-sized data, and RNN representations outperform Transformer representations on semantic similarity tasks.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2019 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Probing Representations Learned by Multimodal Recurrent and Transformer Models
A controlled comparison shows translation and image-grounding objectives produce better sentence representations than language modeling on moderate-sized data, and RNN representations outperform Transformer representations on semantic similarity tasks.