Thinking Fair and Slow: On the Efficacy of Structured Prompts for Debiasing Language Models

Open in new window