2 papers
cs.CL2022
Interpreting BERT-based Text Similarity via Activation and Saliency Maps
Itzik Malkiel, Dvir Ginzburg, Oren Barkan +3
Recently, there has been growing interest in the ability of Transformer-based models to produce meaningful embeddings of text with several applications, such as text similarity. De…
cs.CL2022
MetricBERT: Text Representation Learning via Self-Supervised Triplet Training
Itzik Malkiel, Dvir Ginzburg, Oren Barkan +3
We present MetricBERT, a BERT-based model that learns to embed text under a well-defined similarity metric while simultaneously adhering to the ``traditional'' masked-language task…