no code implementations • 3 Nov 2021 • Dawei Dai, Yutang Li, Huanan Bao, Sy Xia, Guoyin Wang, Xiaoli Ma
From the results, we conclude that (1) the combined effect of certain features is typically far more influential than any single feature; (2) in different tasks, neural models can perform different biases, that is, we can design a specific task to make a neural model biased toward a specific anticipated feature.