Representation biases in sentence transformers
This addresses a gap in understanding representation biases in sentence transformers, which is incremental but important for improving model interpretability and fairness in NLP applications.
The study investigated how sentence transformers produce representations, finding that state-of-the-art models exhibit a strong nominal-participant-set bias, where cosine similarities between sentences are more influenced by noun overlap than by predicates, modifiers, or adjuncts.
Variants of the BERT architecture specialised for producing full-sentence representations often achieve better performance on downstream tasks than sentence embeddings extracted from vanilla BERT. However, there is still little understanding of what properties of inputs determine the properties of such representations. In this study, we construct several sets of sentences with pre-defined lexical and syntactic structures and show that SOTA sentence transformers have a strong nominal-participant-set bias: cosine similarities between pairs of sentences are more strongly determined by the overlap in the set of their noun participants than by having the same predicates, lengthy nominal modifiers, or adjuncts. At the same time, the precise syntactic-thematic functions of the participants are largely irrelevant.