Condenser: A Pre-training Architecture For Dense Retrieval

Luyu Gao, Jamie Callan. Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing 2021 – 76 citations

[Paper]    
Model Architecture Attention Mechanism Transformer Pre-Training Training Techniques

Pre-trained Transformer language models (LM) have become go-to text representation encoders. Prior research fine-tunes deep LMs to encode text sequences such as sentences and passages into single dense vector representations for efficient text comparison and retrieval. However, dense encoders require a lot of data and sophisticated techniques to effectively train and suffer in low data situations. This paper finds a key reason is that standard LMs’ internal attention structure is not ready-to-use for dense encoders, which needs to aggregate text information into the dense representation. We propose to pre-train towards dense encoder with a novel Transformer architecture, Condenser, where LM prediction CONditions on DENSE Representation. Our experiments show Condenser improves over standard LM by large margins on various text retrieval and similarity tasks.

Similar Work