paper-with-me

홈 › Papers

Representation biases in sentence transformers

2023-01-30 · Dmitry Nikolaev, Sebastian Padó

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.

📄 PDF Abstract BibTeX arXiv:2301.13039

Code (0)

등록된 구현이 없습니다.

Tasks

SentenceSentence Embeddings

Methods 이 논문이 사용한 방법론

Refunds@Expedia|||How do I get a full refund from Expedia? “How do I get a full refund from Expedia? How do I get a full refund from Expedia? – Call ☎️ +1-(888) 829 (0881) or +1-805-330-4056 or +1-805-330-4056 for Quick Help &…
Attention 설명 없음
Linear Layer A Linear Layer is a projection $\mathbf{XW + b}$.
Adam 설명 없음
Layer Normalization Unlike batch normalization, Layer Normalization directly estimates the normalization statistics from the summed inputs…
Weight Decay 설명 없음
Multi-Head Attention 설명 없음
Residual Connection 설명 없음

Similar Papers 제목 키워드 기반

Towards Debiasing Sentence Representations

2020-07-16 · ACL 2020 6 · Paul Pu Liang, Irene Mengze Li, Emily Zheng, Yao Chong Lim 외

As natural language processing methods are increasingly deployed in real-world scenarios such as healthcare, legal systems, and social science, it becomes necessary to recognize the role they potentially play in shaping …

Linguistic AcceptabilityNatural Language UnderstandingSentenceSentiment Analysis+1

On Measuring Social Biases in Sentence Encoders

2019-03-25 · NAACL 2019 6 · Chandler May, Alex Wang, Shikha Bordia, Samuel R. Bowman 외

The Word Embedding Association Test shows that GloVe and word2vec word embeddings exhibit human-like implicit biases based on gender, race, and other social constructs (Caliskan et al., 2017). Meanwhile, research on lear…

SentenceWord Embeddings

Transformer Grammars: Augmenting Transformer Language Models with Syntactic Inductive Biases at Scale

2022-03-01 · Laurent Sartran, Samuel Barrett, Adhiguna Kuncoro, Miloš Stanojević 외

We introduce Transformer Grammars (TGs), a novel class of Transformer language models that combine (i) the expressive power, scalability, and strong performance of Transformers and (ii) recursive syntactic compositions, …

Inductive BiasLanguage ModelingLanguage ModellingSentence

Information Representation Fairness in Long-Document Embeddings: The Peculiar Interaction of Positional and Language Bias

2026-01-23 · Elias Schuhmacher, Andrianos Michail, Juri Opitz, Rico Sennrich 외 arxiv

To be discoverable in an embedding-based search process, each part of a document should be reflected in its embedding representation. To quantify any potential reflection biases, we introduce a permutation-based evaluati…

Deep Representational Re-tuning using Contrastive Tension

2021-01-01 · ICLR 2021 1 · Fredrik Carlsson, Amaru Cuba Gyllensten, Evangelia Gogoulou, Erik Ylipää Hellqvist 외

Extracting semantically useful natural language sentence representations from pre-trained deep neural networks such as Transformers remains a challenge. We first demonstrate that pre-training objectives impose a signific…

Semantic SimilaritySemantic Textual SimilaritySentenceSentence Embedding+2