# Representation biases in sentence transformers

arXiv: https://arxiv.org/abs/2301.13039 · HTML: https://mlsanity.com/paper/2301.13039/

## Key facts

- Cited by 7 papers in the map (data through 14 Oct 2026).
- Builds on 20 of its 35 references that are in the map.
- First posted on arXiv 30 Jan 2023.

## Builds on (most cited first, 20 of 20)

- [BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding](https://mlsanity.com/paper/1810.04805/) (2018)
- [RoBERTa: A Robustly Optimized BERT Pretraining Approach](https://mlsanity.com/paper/1907.11692/) (2019)
- [Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks](https://mlsanity.com/paper/1908.10084/) (2019)
- [DistilBERT, a distilled version of BERT: smaller, faster, cheaper and lighter](https://mlsanity.com/paper/1910.01108/) (2019)
- [UMAP: Uniform Manifold Approximation and Projection for Dimension Reduction](https://mlsanity.com/paper/1802.03426/) (2018)
- [BERT Rediscovers the Classical NLP Pipeline](https://mlsanity.com/paper/1905.05950/) (2019)
- [MPNet: Masked and Permuted Pre-training for Language Understanding](https://mlsanity.com/paper/2004.09297/) (2020)
- [A Primer in BERTology: What we know about how BERT works](https://mlsanity.com/paper/2002.12327/) (2020)
- [Targeted Syntactic Evaluation of Language Models](https://mlsanity.com/paper/1808.09031/) (2018)
- [Examining Gender and Race Bias in Two Hundred Sentiment Analysis Systems](https://mlsanity.com/paper/1805.04508/) (2018)
- [Deep Learning Based Text Classification: A Comprehensive Review](https://mlsanity.com/paper/2004.03705/) (2020)
- [On the Sentence Embeddings from Pre-trained Language Models](https://mlsanity.com/paper/2011.05864/) (2020)
- [Improving Named Entity Recognition by External Context Retrieving and Cooperative Learning](https://mlsanity.com/paper/2105.03654/) (2021)
- [Positional Artefacts Propagate Through Masked Language Model Embeddings](https://mlsanity.com/paper/2011.04393/) (2020)
- [SBERT studies Meaning Representations: Decomposing Sentence Embeddings into Explainable Semantic Features](https://mlsanity.com/paper/2206.07023/) (2022)
- [Universal Text Representation from BERT: An Empirical Study](https://mlsanity.com/paper/1910.07973/) (2019)
- [ABNIRML: Analyzing the Behavior of Neural IR Models](https://mlsanity.com/paper/2011.00696/) (2020)
- [Laughing Heads: Can Transformers Detect What Makes a Sentence Funny?](https://mlsanity.com/paper/2105.09142/) (2021)
- [Don't Settle for Average, Go for the Max: Fuzzy Sets and Max-Pooled Word Vectors](https://mlsanity.com/paper/1904.13264/) (2019)
- [SyGNS: A Systematic Generalization Testbed Based on Natural Language Semantics](https://mlsanity.com/paper/2106.01077/) (2021)

## Built on by (most cited first, 7 of 7)

- [Steering Dialogue Dynamics for Robustness against Multi-turn Jailbreaking Attacks](https://mlsanity.com/paper/2503.00187/) (2025)
- [An Attribution Method for Siamese Encoders](https://mlsanity.com/paper/2310.05703/) (2023)
- [Grammatical information in BERT sentence embeddings as two-dimensional arrays](https://mlsanity.com/paper/2312.09890/) (2023)
- [Adverbs, Surprisingly](https://mlsanity.com/paper/2305.19650/) (2023)
- [BERT-Enhanced Retrieval Tool for Homework Plagiarism Detection System](https://mlsanity.com/paper/2404.01582/) (2024)
- [Distance Is All You Need: Radial Dispersion for Uncertainty Estimation in Large Language Models](https://mlsanity.com/paper/2512.04351/) (2025)
- [Zero-Shot Warning Generation for Misinformative Multimodal Content](https://mlsanity.com/paper/2502.00752/) (2025)

## Abstract (arXiv)

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.