Attacks against Abstractive Text Summarization Models through Lead Bias and Influence Functions.

Poojitha Thota, Shirin Nilizadeh

Browse the full EMNLP paper archive.