Diverse Adversaries for Mitigating Bias in Training

TL;DR

This paper introduces a novel adversarial learning method using multiple diverse discriminators with orthogonal representations, significantly improving bias mitigation and training stability in language models.

Contribution

The paper proposes a new adversarial training approach with multiple diverse discriminators to better mitigate bias and enhance training stability.

Findings

Substantially reduces bias compared to standard methods

Improves training stability in adversarial learning

Demonstrates effectiveness on language models

Abstract

Adversarial learning can learn fairer and less biased models of language than standard methods. However, current adversarial techniques only partially mitigate model bias, added to which their training procedures are often unstable. In this paper, we propose a novel approach to adversarial learning based on the use of multiple diverse discriminators, whereby discriminators are encouraged to learn orthogonal hidden representations from one another. Experimental results show that our method substantially improves over standard adversarial removal methods, in terms of reducing bias and the stability of training.