# How Gender Debiasing Affects Internal Model Representations, and Why It Matters

arXiv: https://arxiv.org/abs/2204.06827 · HTML: https://mlsanity.com/paper/2204.06827/

## Key facts

- Cited by 12 papers in the map (data through 14 Oct 2026).
- Builds on 37 of its 58 references that are in the map.
- First posted on arXiv 14 Apr 2022.

## Builds on (most cited first, 30 of 37)

- [PyTorch: An Imperative Style, High-Performance Deep Learning Library](https://mlsanity.com/paper/1912.01703/) (2019)
- [RoBERTa: A Robustly Optimized BERT Pretraining Approach](https://mlsanity.com/paper/1907.11692/) (2019)
- [On the Opportunities and Risks of Foundation Models](https://mlsanity.com/paper/2108.07258/) (2021)
- [DeBERTa: Decoding-enhanced BERT with Disentangled Attention](https://mlsanity.com/paper/2006.03654/) (2020)
- [Man is to Computer Programmer as Woman is to Homemaker? Debiasing Word Embeddings](https://mlsanity.com/paper/1607.06520/) (2016)
- [Language (Technology) is Power: A Critical Survey of "Bias" in NLP](https://mlsanity.com/paper/2005.14050/) (2020)
- [Semantics derived automatically from language corpora contain human-like biases](https://mlsanity.com/paper/1608.07187/) (2016)
- [Gender Bias in Coreference Resolution: Evaluation and Debiasing Methods](https://mlsanity.com/paper/1804.06876/) (2018)
- [Designing and Interpreting Probes with Control Tasks](https://mlsanity.com/paper/1909.03368/) (2019)
- [Null It Out: Guarding Protected Attributes by Iterative Nullspace Projection](https://mlsanity.com/paper/2004.07667/) (2020)
- [Gender Bias in Coreference Resolution](https://mlsanity.com/paper/1804.09301/) (2018)
- [BOLD: Dataset and Metrics for Measuring Biases in Open-Ended Language Generation](https://mlsanity.com/paper/2101.11718/) (2021)
- [On Measuring Social Biases in Sentence Encoders](https://mlsanity.com/paper/1903.10561/) (2019)
- [Bias in Bios: A Case Study of Semantic Representation Bias in a High-Stakes Setting](https://mlsanity.com/paper/1901.09451/) (2019)
- [Analysis Methods in Neural Language Processing: A Survey](https://mlsanity.com/paper/1812.08951/) (2018)
- [Measuring Bias in Contextualized Word Representations](https://mlsanity.com/paper/1906.07337/) (2019)
- [Lipstick on a Pig: Debiasing Methods Cover up Systematic Gender Biases in Word Embeddings But do not Remove Them](https://mlsanity.com/paper/1903.03862/) (2019)
- [Evaluating Gender Bias in Machine Translation](https://mlsanity.com/paper/1906.00591/) (2019)
- [Examining Gender and Race Bias in Two Hundred Sentiment Analysis Systems](https://mlsanity.com/paper/1805.04508/) (2018)
- [Information-Theoretic Probing with Minimum Description Length](https://mlsanity.com/paper/2003.12298/) (2020)
- [Gender Bias in Contextualized Word Embeddings](https://mlsanity.com/paper/1904.03310/) (2019)
- [A Framework for Understanding Sources of Harm throughout the Machine Learning Life Cycle](https://mlsanity.com/paper/1901.10002/) (2019)
- [Detecting Emergent Intersectional Biases: Contextualized Word Embeddings Contain a Distribution of Human-like Biases](https://mlsanity.com/paper/2006.03955/) (2020)
- [Higher-order Coreference Resolution with Coarse-to-fine Inference](https://mlsanity.com/paper/1804.05392/) (2018)
- [Intrinsic Bias Metrics Do Not Correlate with Application Bias](https://mlsanity.com/paper/2012.15859/) (2020)
- [Assessing Social and Intersectional Biases in Contextualized Word Representations](https://mlsanity.com/paper/1911.01485/) (2019)
- [Evaluating the Underlying Gender Bias in Contextualized Word Embeddings](https://mlsanity.com/paper/1904.08783/) (2019)
- [Toward Gender-Inclusive Coreference Resolution](https://mlsanity.com/paper/1910.13913/) (2019)
- [On the Intrinsic and Extrinsic Fairness Evaluation Metrics for Contextualized Language Representations](https://mlsanity.com/paper/2203.13928/) (2022)
- [Understanding Undesirable Word Embedding Associations](https://mlsanity.com/paper/1908.06361/) (2019)

## Built on by (most cited first, 12 of 12)

- [Bias and Fairness in Large Language Models: A Survey](https://mlsanity.com/paper/2309.00770/) (2023)
- [MABEL: Attenuating Gender Bias using Textual Entailment Data](https://mlsanity.com/paper/2210.14975/) (2022)
- [Choose Your Lenses: Flaws in Gender Bias Evaluation](https://mlsanity.com/paper/2210.11471/) (2022)
- [Shielded Representations: Protecting Sensitive Attributes Through Iterative Gradient-Based Projection](https://mlsanity.com/paper/2305.10204/) (2023)
- [On Behalf of the Stakeholders: Trends in NLP Model Interpretability in the Era of LLMs](https://mlsanity.com/paper/2407.19200/) (2024)
- [Investigating Implicit Bias in Large Language Models: A Large-Scale Study of Over 50 LLMs](https://mlsanity.com/paper/2410.12864/) (2024)
- [BLIND: Bias Removal With No Demographics](https://mlsanity.com/paper/2212.10563/) (2022)
- [Measures of Information Reflect Memorization Patterns](https://mlsanity.com/paper/2210.09404/) (2022)
- [How Far Can It Go?: On Intrinsic Gender Bias Mitigation for Text Classification](https://mlsanity.com/paper/2301.12855/) (2023)
- [Cross-lingual Transfer Can Worsen Bias in Sentiment Analysis](https://mlsanity.com/paper/2305.12709/) (2023)
- [Evaluating Gender Bias Transfer between Pre-trained and Prompt-Adapted Language Models](https://mlsanity.com/paper/2412.03537/) (2024)
- [Log-linear Guardedness and its Implications](https://mlsanity.com/paper/2210.10012/) (2022)

## Abstract (arXiv)

Common studies of gender bias in NLP focus either on extrinsic bias measured by model performance on a downstream task or on intrinsic bias found in models' internal representations. However, the relationship between extrinsic and intrinsic bias is relatively unknown. In this work, we illuminate this relationship by measuring both quantities together: we debias a model during downstream fine-tuning, which reduces extrinsic bias, and measure the effect on intrinsic bias, which is operationalized as bias extractability with information-theoretic probing. Through experiments on two tasks and multiple bias metrics, we show that our intrinsic bias metric is a better indicator of debiasing than (a contextual adaptation of) the standard WEAT metric, and can also expose cases of superficial debiasing. Our framework provides a comprehensive perspective on bias in NLP models, which can be applied to deploy NLP systems in a more informed manner. Our code and model checkpoints are publicly available.