# Double-Hard Debias: Tailoring Word Embeddings for Gender Bias Mitigation

arXiv: https://arxiv.org/abs/2005.00965 · HTML: https://mlsanity.com/paper/2005.00965/

## Key facts

- Cited by 36 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 3 May 2020.

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

- [Sparse Feature Circuits: Discovering and Editing Interpretable Causal Graphs in Language Models](https://mlsanity.com/paper/2403.19647/) (2024)
- [Mitigating Language-Dependent Ethnic Bias in BERT](https://mlsanity.com/paper/2109.05704/) (2021)
- [Theories of "Gender" in NLP Bias Research](https://mlsanity.com/paper/2205.02526/) (2022)
- [Low Frequency Names Exhibit Bias and Overfitting in Contextualizing Language Models](https://mlsanity.com/paper/2110.00672/) (2021)
- [Steering Out-of-Distribution Generalization with Concept Ablation Fine-Tuning](https://mlsanity.com/paper/2507.16795/) (2025)
- [Shielded Representations: Protecting Sensitive Attributes Through Iterative Gradient-Based Projection](https://mlsanity.com/paper/2305.10204/) (2023)
- [Gender Bias in Word Embeddings: A Comprehensive Analysis of Frequency, Syntax, and Semantics](https://mlsanity.com/paper/2206.03390/) (2022)
- [Debiasing Methods for Fairer Neural Models in Vision and Language Research: A Survey](https://mlsanity.com/paper/2211.05617/) (2022)
- [Does Robustness Improve Fairness? Approaching Fairness with Word Substitution Robustness Methods for Text Classification](https://mlsanity.com/paper/2106.10826/) (2021)
- [Sense Embeddings are also Biased--Evaluating Social Biases in Static and Contextualised Sense Embeddings](https://mlsanity.com/paper/2203.07523/) (2022)
- [The Birth of Bias: A case study on the evolution of gender bias in an English language model](https://mlsanity.com/paper/2207.10245/) (2022)
- [Learning to Generate Equitable Text in Dialogue from Biased Training Data](https://mlsanity.com/paper/2307.04303/) (2023)
- [Bias Vector: Mitigating Biases in Language Models with Task Arithmetic Approach](https://mlsanity.com/paper/2412.11679/) (2024)
- [Word Embeddings via Causal Inference: Gender Bias Reducing and Semantic Information Preserving](https://mlsanity.com/paper/2112.05194/) (2021)
- [Large Human Language Models: A Need and the Challenges](https://mlsanity.com/paper/2312.07751/) (2023)
- [Position: LLMs Must Use Functor-Based and RAG-Driven Bias Mitigation for Fairness](https://mlsanity.com/paper/2603.07368/) (2026)
- [Fair NLP Models with Differentially Private Text Encoders](https://mlsanity.com/paper/2205.06135/) (2022)
- [CatRAG: Functor-Guided Structural Debiasing with Retrieval Augmentation for Fair LLMs](https://mlsanity.com/paper/2603.21524/) (2026)
- [WordBias: An Interactive Visual Tool for Discovering Intersectional Biases Encoded in Word Embeddings](https://mlsanity.com/paper/2103.03598/) (2021)
- [Marked Attribute Bias in Natural Language Inference](https://mlsanity.com/paper/2109.14039/) (2021)
- [Estimating and Maximizing Mutual Information for Knowledge Distillation](https://mlsanity.com/paper/2110.15946/) (2021)
- [Gender Bias in Meta-Embeddings](https://mlsanity.com/paper/2205.09867/) (2022)
- [Attention Pruning: Automated Fairness Repair of Language Models via Surrogate Simulated Annealing](https://mlsanity.com/paper/2503.15815/) (2025)
- [Fair Text Classification with Wasserstein Independence](https://mlsanity.com/paper/2311.12689/) (2023)
- [On the Mutual Influence of Gender and Occupation in LLM Representations](https://mlsanity.com/paper/2503.06792/) (2025)
- [Chain-of-Thought Driven Adversarial Scenario Extrapolation for Robust Language Models](https://mlsanity.com/paper/2505.17089/) (2025)
- [VERB: Visualizing and Interpreting Bias Mitigation Techniques for Word Representations](https://mlsanity.com/paper/2104.02797/) (2021)
- [Algorithmic Fairness: A Tolerance Perspective](https://mlsanity.com/paper/2405.09543/) (2024)
- [Preliminary Steps Towards Federated Sentiment Classification](https://mlsanity.com/paper/2107.11956/) (2021)
- [Adversarial Examples Generation for Reducing Implicit Gender Bias in Pre-trained Models](https://mlsanity.com/paper/2110.01094/) (2021)

## Abstract (arXiv)

Word embeddings derived from human-generated corpora inherit strong gender bias which can be further amplified by downstream models. Some commonly adopted debiasing approaches, including the seminal Hard Debias algorithm, apply post-processing procedures that project pre-trained word embeddings into a subspace orthogonal to an inferred gender subspace. We discover that semantic-agnostic corpus regularities such as word frequency captured by the word embeddings negatively impact the performance of these algorithms. We propose a simple but effective technique, Double Hard Debias, which purifies the word embeddings against such corpus regularities prior to inferring and removing the gender subspace. Experiments on three bias mitigation benchmarks show that our approach preserves the distributional semantics of the pre-trained word embeddings while reducing gender bias to a significantly larger degree than prior approaches.