# The Impact of Debiasing on the Performance of Language Models in Downstream Tasks is Underestimated

arXiv: https://arxiv.org/abs/2309.09092 · HTML: https://mlsanity.com/paper/2309.09092/

## Key facts

- Cited by 7 papers in the map (data through 14 Oct 2026).
- Builds on 23 of its 35 references that are in the map.
- First posted on arXiv 16 Sep 2023.

## Builds on (most cited first, 23 of 23)

- [BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding](https://mlsanity.com/paper/1810.04805/) (2018)
- [RoBERTa: A Robustly Optimized BERT Pretraining Approach](https://mlsanity.com/paper/1907.11692/) (2019)
- [GLUE: A Multi-Task Benchmark and Analysis Platform for Natural Language Understanding](https://mlsanity.com/paper/1804.07461/) (2018)
- [SQuAD: 100,000+ Questions for Machine Comprehension of Text](https://mlsanity.com/paper/1606.05250/) (2016)
- [A Broad-Coverage Challenge Corpus for Sentence Understanding through Inference](https://mlsanity.com/paper/1704.05426/) (2017)
- [Man is to Computer Programmer as Woman is to Homemaker? Debiasing Word Embeddings](https://mlsanity.com/paper/1607.06520/) (2016)
- [SemEval-2017 Task 1: Semantic Textual Similarity - Multilingual and Cross-lingual Focused Evaluation](https://mlsanity.com/paper/1708.00055/) (2017)
- [Neural Network Acceptability Judgments](https://mlsanity.com/paper/1805.12471/) (2018)
- [StereoSet: Measuring stereotypical bias in pretrained language models](https://mlsanity.com/paper/2004.09456/) (2020)
- [CrowS-Pairs: A Challenge Dataset for Measuring Social Biases in Masked Language Models](https://mlsanity.com/paper/2010.00133/) (2020)
- [Measuring Bias in Contextualized Word Representations](https://mlsanity.com/paper/1906.07337/) (2019)
- [Measuring and Reducing Gendered Correlations in Pre-trained Models](https://mlsanity.com/paper/2010.06032/) (2020)
- [Gender Bias in Contextualized Word Embeddings](https://mlsanity.com/paper/1904.03310/) (2019)
- [Counterfactual Data Augmentation for Mitigating Gender Stereotypes in Languages with Rich Morphology](https://mlsanity.com/paper/1906.04571/) (2019)
- [Harms of Gender Exclusivity and Challenges in Non-Binary Representation in Language Technologies](https://mlsanity.com/paper/2108.12084/) (2021)
- [On Measuring and Mitigating Biased Inferences of Word Embeddings](https://mlsanity.com/paper/1908.09369/) (2019)
- [Debiasing Pre-trained Contextualised Embeddings](https://mlsanity.com/paper/2101.09523/) (2021)
- [An Empirical Survey of the Effectiveness of Debiasing Techniques for Pre-trained Language Models](https://mlsanity.com/paper/2110.08527/) (2021)
- [Toward Gender-Inclusive Coreference Resolution](https://mlsanity.com/paper/1910.13913/) (2019)
- [Unmasking the Mask -- Evaluating Social Biases in Masked Language Models](https://mlsanity.com/paper/2104.07496/) (2021)
- [Gender Bias in Masked Language Models for Multiple Languages](https://mlsanity.com/paper/2205.00551/) (2022)
- [Sense Embeddings are also Biased--Evaluating Social Biases in Static and Contextualised Sense Embeddings](https://mlsanity.com/paper/2203.07523/) (2022)
- [Comparing Intrinsic Gender Bias Evaluation Measures without using Human Annotated Examples](https://mlsanity.com/paper/2301.12074/) (2023)

## Built on by (most cited first, 7 of 7)

- [Dissecting Bias in LLMs: A Mechanistic Interpretability Perspective](https://mlsanity.com/paper/2506.05166/) (2025)
- [Likelihood-based Mitigation of Evaluation Bias in Large Language Models](https://mlsanity.com/paper/2402.15987/) (2024)
- [The Gaps between Pre-train and Downstream Settings in Bias Evaluation and Debiasing](https://mlsanity.com/paper/2401.08511/) (2024)
- [Bias in, Bias out: Annotation Bias in Multilingual Large Language Models](https://mlsanity.com/paper/2511.14662/) (2025)
- [Bias Mitigation or Cultural Commonsense? Evaluating LLMs with a Japanese Dataset](https://mlsanity.com/paper/2509.24468/) (2025)
- [Fair-GPTQ: Bias-Aware Quantization for Large Language Models](https://mlsanity.com/paper/2509.15206/) (2025)
- [A Representation-Level Assessment of Bias Mitigation in Foundation Models](https://mlsanity.com/paper/2604.08561/) (2026)

## Abstract (arXiv)

Pre-trained language models trained on large-scale data have learned serious levels of social biases. Consequently, various methods have been proposed to debias pre-trained models. Debiasing methods need to mitigate only discriminatory bias information from the pre-trained models, while retaining information that is useful for the downstream tasks. In previous research, whether useful information is retained has been confirmed by the performance of downstream tasks in debiased pre-trained models. On the other hand, it is not clear whether these benchmarks consist of data pertaining to social biases and are appropriate for investigating the impact of debiasing. For example in gender-related social biases, data containing female words (e.g. ``she, female, woman''), male words (e.g. ``he, male, man''), and stereotypical words (e.g. ``nurse, doctor, professor'') are considered to be the most affected by debiasing. If there is not much data containing these words in a benchmark dataset for a target task, there is the possibility of erroneously evaluating the effects of debiasing. In this study, we compare the impact of debiasing on performance across multiple downstream tasks using a wide-range of benchmark datasets that containing female, male, and stereotypical words. Experiments show that the effects of debiasing are consistently underestimated across all tasks. Moreover, the effects of debiasing could be reliably evaluated by separately considering instances containing female, male, and stereotypical words than all of the instances in a benchmark dataset.