Perturbation Augmentation for Fairer NLP

2022/05/25 by Rebecca Qian, Qian, Rebecca, Candace Ross +9 · 4 citations
Computer Science · #Artificial Intelligence (cs.AI) #Computation and Language (cs.CL) #FOS: Computer and information sciences #Topic Modeling

paper · pdf · doi:10.48550/arxiv.2205.12586

openalex publication_date 2022/05/25 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

Unwanted and often harmful social biases are becoming ever more salient in NLP research, affecting both models and datasets. In this work, we ask whether training on demographically perturbed data leads to fairer language models. We collect a large dataset of human annotated text perturbations and train a neural perturbation model, which we show outperforms heuristic alternatives. We find that (i) language models (LMs) pre-trained on demographically perturbed corpora are typically more fair, and (ii) LMs finetuned on perturbed GLUE datasets exhibit less demographic bias on downstream tasks, and (iii) fairness improvements do not come at the expense of performance on downstream tasks. Lastly, we discuss outstanding questions about how best to evaluate the (un)fairness of large language models. We hope that this exploration of neural demographic perturbation will help drive more improvement towards fairer NLP.

Cited by

• Investigating Intersectional Bias in Large Language Models using Confidence Disparities in Coreference Resolution
• Embodied AI Agents: Modeling the World
• Robustness in Large Language Models: A Survey of Mitigation Strategies and Evaluation Metrics
• Gender Bias in Explainability: Investigating Performance Disparity in Post-hoc Methods

Related

• Fairness Pruning: Locating Demographic Bias in GLU-MLP Layers via Differential Activations
• Bias and Fairness in Large Language Models: A Survey
• A Survey on Bias and Fairness in Natural Language Processing
• Bias and Fairness in Large Language Models: A Survey
• Attention Pruning: Automated Fairness Repair of Language Models via Surrogate Simulated Annealing