Bias Vector: Mitigating Biases in Language Models with Task Arithmetic Approach.

Daiki Shirafuji, Makoto Takenaka, Shinya Taguchi

Browse the full COLING paper archive.