Mitigation of biases, such as language models’ reliance on gender stereotypes, is a crucial endeavor required for the creation of reliable and useful language technology. The crucial aspect of debiasing is to ensure that the models preserve their versatile capabilities, including their ability to solve language tasks and equitably represent various genders. To address these issues, we introduce Dual Dabiasing Algorithm through Model Adaptation (2DAMA). Novel Dual Debiasing enables robust reduction of stereotypical bias while preserving desired factual gender information encoded by language models. We show that 2DAMA effectively reduces gender bias in language models for English and is one of the first approaches facilitating the mitigation of their stereotypical tendencies in translation. The proposed method’s key advantage is the preservation of factual gender cues, which are useful in a wide range of natural language processing tasks.

Independent Component Analysis (ICA) is an algorithm originally developed for finding separate sources in a mixed signal, such as a recording of multiple people in the same room speaking at the same time. Unlike Principal Component Analysis (PCA), ICA permits the representation of a word as an unstructured set of features, without any particular feature being deemed more significant than the others. In this paper, we used ICA to analyze word embeddings. We have found that ICA can be used to find semantic features of the words and these features can easily be combined to search for words that satisfy the combination. We show that most of the independent components represent such features. To quantify the interpretability of the components, we use the word intruder test, performed both by humans and by large language models. We propose to use the automated version of the word intruder test as a fast and inexpensive way of quantifying vector interpretability without the need for human effort.

Large language models are becoming the go-to solution for the ever-growing number of tasks. However, with growing capacity, models are prone to rely on spurious correlations stemming from biases and stereotypes present in the training data. This work proposes a novel method for detecting and mitigating gender bias in language models. We perform causal analysis to identify problematic model components and discover that mid-upper feed-forward layers are most prone to convey bias. Based on the analysis results, we intervene in the model by applying a linear projection to the weight matrices of these layers. Our titular method DAMA, significantly decreases bias as measured by diverse metrics while maintaining the model's performance on downstream tasks. We release code for our method and models, which retrain LLaMA's state-of-the-art performance while being significantly less biased.

Recent studies have shown that transformer models like BERT rely on number information encoded in their representations of sentences' subjects and head verbs when performing subject-verb agreement.However, probing experiments suggest that subject number is also encoded in the representations of all words in such sentences.In this paper, we use causal interventions to show that BERT only uses the subject plurality information encoded in its representations of the subject and words that agree with it in number.We also demonstrate that current probing metrics are unable to determine which words' representations contain functionally relevant information.This both provides a revised view of subject-verb agreement in language models, and suggests potential pitfalls for current probe usage and evaluation.

Bar Iluz, Tomasz Limisiewicz, Gabriel Stanovsky, David Mareček. Proceedings of the 13th International Joint Conference on Natural Language Processing and the 3rd Conference of the Asia-Pacific Chapter of the Association for Computational Linguistics (Volume 1: Long Papers). 2023.

Tomasz Limisiewicz, David Mareček. Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers). 2021.

State-of-the-art contextual embeddings are obtained from large language models available only for a few languages.For others, we need to learn representations using a multilingual model.There is an ongoing debate on whether multilingual embeddings can be aligned in a space shared across many languages.The novel Orthogonal Structural Probe (Limisiewicz and Mareček, 2021) allows us to answer this question for specific linguistic features and learn a projection based only on mono-lingual annotated datasets.We evaluate syntactic (UD) and lexical (WordNet) structural information encoded in MBERT's contextual representations for nine diverse languages.1 We observe that for languages closely related to English, no transformation is needed.The evaluated information is encoded in a shared cross-lingual embedding space.For other languages, it is beneficial to apply orthogonal transformation learned separately for each language.We successfully apply our findings to zero-shot and few-shot cross-lingual parsing.