An Empirical Survey of the Effectiveness of Debiasing Techniques for Pre-trained Language Models

2021/10/16 by Nicholas Meade, Elinor Poole-Dayan, Meade, Nicholas +3 · 5 citations
Computer Science · #Computation and Language (cs.CL) #FOS: Computer and information sciences #Machine Learning (cs.LG) #Natural Language Processing Techniques #Topic Modeling

paper · pdf · doi:10.48550/arxiv.2110.08527

openalex publication_date 2021/10/16 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

Recent work has shown pre-trained language models capture social biases from the large amounts of text they are trained on. This has attracted attention to developing techniques that mitigate such biases. In this work, we perform an empirical survey of five recently proposed bias mitigation techniques: Counterfactual Data Augmentation (CDA), Dropout, Iterative Nullspace Projection, Self-Debias, and SentenceDebias. We quantify the effectiveness of each technique using three intrinsic bias benchmarks while also measuring the impact of these techniques on a model's language modeling ability, as well as its performance on downstream NLU tasks. We experimentally find that: (1) Self-Debias is the strongest debiasing technique, obtaining improved scores on all bias benchmarks; (2) Current debiasing techniques perform less consistently when mitigating non-gender biases; And (3) improvements on bias benchmarks such as StereoSet and CrowS-Pairs by using debiasing strategies are often accompanied by a decrease in language modeling ability, making it difficult to determine whether the bias mitigation was effective.

Cited by

• Large Language Models Develop Novel Social Biases Through Adaptive Exploration
• TriCon-Fair: Triplet Contrastive Learning for Mitigating Social Bias in Pre-trained Language Models
• BiasFreeBench: a Benchmark for Mitigating Bias in Large Language Model Responses
• Fair-GPTQ: Bias-Aware Quantization for Large Language Models
• PromptGuard: An Orchestrated Prompting Framework for Principled Synthetic Text Generation for Vulnerable Populations using LLMs with Enhanced Safety, Fairness, and Controllability

Related

• The Impact of Debiasing on the Performance of Language Models in Downstream Tasks is Underestimated
• An Empirical Analysis of Parameter-Efficient Methods for Debiasing Pre-Trained Language Models
• Mitigating Biases in Language Models via Bias Unlearning
• Debiasing Algorithm through Model Adaptation
• Prompt Tuning Pushes Farther, Contrastive Learning Pulls Closer: A Two-Stage Approach to Mitigate Social Biases