TY  - RPRT
TI  - Learning Optimal Representations with the Decodable Information Bottleneck
AU  - Yann Dubois
AU  - Douwe Kiela
AU  - David J. Schwab
AU  - Ramakrishna Vedantam
PY  - 2021
UR  - https://arxiv.org/abs/2009.12789
ID  - 2009.12789
ER  -