# A General Framework for Implicit and Explicit Debiasing of Distributional Word Vector Spaces

arXiv: https://arxiv.org/abs/1909.06092 · HTML: https://mlsanity.com/paper/1909.06092/

## Key facts

- Cited by 36 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 13 Sep 2019.

## Built on by (most cited first, 30 of 36)

- [RedditBias: A Real-World Resource for Bias Evaluation and Debiasing of Conversational Language Models](https://mlsanity.com/paper/2106.03521/) (2021)
- [Sustainable Modular Debiasing of Language Models](https://mlsanity.com/paper/2109.03646/) (2021)
- [Theories of "Gender" in NLP Bias Research](https://mlsanity.com/paper/2205.02526/) (2022)
- [Welcome to the Modern World of Pronouns: Identity-Inclusive Natural Language Processing beyond Gender](https://mlsanity.com/paper/2202.11923/) (2022)
- [Model Editing as a Robust and Denoised variant of DPO: A Case Study on Toxicity](https://mlsanity.com/paper/2405.13967/) (2024)
- [CHBias: Bias Evaluation and Mitigation of Chinese Conversational Language Models](https://mlsanity.com/paper/2305.11262/) (2023)
- [Sociodemographic Bias in Language Models: A Survey and Forward Path](https://mlsanity.com/paper/2306.08158/) (2023)
- [Zero-Shot Robustification of Zero-Shot Models](https://mlsanity.com/paper/2309.04344/) (2023)
- [Debiasing Methods for Fairer Neural Models in Vision and Language Research: A Survey](https://mlsanity.com/paper/2211.05617/) (2022)
- [Bias and Fairness in Chatbots: An Overview](https://mlsanity.com/paper/2309.08836/) (2023)
- [What about em? How Commercial Machine Translation Fails to Handle (Neo-)Pronouns](https://mlsanity.com/paper/2305.16051/) (2023)
- [Fair and Argumentative Language Modeling for Computational Argumentation](https://mlsanity.com/paper/2204.04026/) (2022)
- [Bias and Volatility: A Statistical Framework for Evaluating Large Language Model's Stereotypes and the Associated Generation Inconsistency](https://mlsanity.com/paper/2402.15481/) (2024)
- [Diachronic Analysis of German Parliamentary Proceedings: Ideological Shifts through the Lens of Political Biases](https://mlsanity.com/paper/2108.06295/) (2021)
- [On the Emergence of Linear Analogies in Word Embeddings](https://mlsanity.com/paper/2505.18651/) (2025)
- [Position: LLMs Must Use Functor-Based and RAG-Driven Bias Mitigation for Fairness](https://mlsanity.com/paper/2603.07368/) (2026)
- [Closed-Form Training Dynamics Reveal Learned Features and Linear Structure in Word2Vec-like Models](https://mlsanity.com/paper/2502.09863/) (2025)
- [Marked Attribute Bias in Natural Language Inference](https://mlsanity.com/paper/2109.14039/) (2021)
- [Measuring and Addressing Indexical Bias in Information Retrieval](https://mlsanity.com/paper/2406.04298/) (2024)
- [DebIE: A Platform for Implicit and Explicit Debiasing of Word Embedding Spaces](https://mlsanity.com/paper/2103.06598/) (2021)
- [Bridging Fairness and Environmental Sustainability in Natural Language Processing](https://mlsanity.com/paper/2211.04256/) (2022)
- [A Group Fairness Lens for Large Language Models](https://mlsanity.com/paper/2312.15478/) (2023)
- [Language Invariant Properties in Natural Language Processing](https://mlsanity.com/paper/2109.13037/) (2021)
- [Back to the Future: On Potential Histories in NLP](https://mlsanity.com/paper/2210.06245/) (2022)
- [COBIAS: Assessing the Contextual Reliability of Bias Benchmarks for Language Models](https://mlsanity.com/paper/2402.14889/) (2024)
- [Local Contrastive Editing of Gender Stereotypes](https://mlsanity.com/paper/2410.17739/) (2024)
- [Speciesist Language and Nonhuman Animal Bias in English Masked Language Models](https://mlsanity.com/paper/2203.05140/) (2022)
- [Does Debiasing Inevitably Degrade the Model Performance](https://mlsanity.com/paper/2211.07350/) (2022)
- [AI Fairness: from Principles to Practice](https://mlsanity.com/paper/2207.09833/) (2022)
- [For the Underrepresented in Gender Bias Research: Chinese Name Gender Prediction with Heterogeneous Graph Attention Network](https://mlsanity.com/paper/2302.00419/) (2023)

## Abstract (arXiv)

Distributional word vectors have recently been shown to encode many of the human biases, most notably gender and racial biases, and models for attenuating such biases have consequently been proposed. However, existing models and studies (1) operate on under-specified and mutually differing bias definitions, (2) are tailored for a particular bias (e.g., gender bias) and (3) have been evaluated inconsistently and non-rigorously. In this work, we introduce a general framework for debiasing word embeddings. We operationalize the definition of a bias by discerning two types of bias specification: explicit and implicit. We then propose three debiasing models that operate on explicit or implicit bias specifications and that can be composed towards more robust debiasing. Finally, we devise a full-fledged evaluation framework in which we couple existing bias metrics with newly proposed ones. Experimental findings across three embedding methods suggest that the proposed debiasing models are robust and widely applicable: they often completely remove the bias both implicitly and explicitly without degradation of semantic information encoded in any of the input distributional spaces. Moreover, we successfully transfer debiasing models, by means of cross-lingual embedding spaces, and remove or attenuate biases in distributional word vector spaces of languages that lack readily available bias specifications.