KLAAD: Refining Attention Mechanisms to Reduce Societal Bias in Generative Language Models.

Seorin Kim, Dongyoung Lee, Jaejin Lee

Browse the full EMNLP paper archive.