Upstream Mitigation Is Not All You Need: Testing the Bias Transfer Hypothesis in Pre-Trained Language Models

Presentation @ ACL 2022

• View the slides here.
• Read the paper here.