Task-Attentive Transformer Architecture for Continual Learning of Vision-and-Language Tasks Using Knowledge Distillation.

Yuliang Cai, Jesse Thomason, Mohammad Rostami

Browse the full EMNLP paper archive.