Interpreting Pretrained Language Models via Concept Bottlenecks.

Zhen Tan, Lu Cheng, Song Wang, Bo Yuan, Jundong Li, Huan Liu

Browse the full PAKDD paper archive.