all AI news
A Song of (Dis)agreement: Evaluating the Evaluation of Explainable Artificial Intelligence in Natural Language Processing. (arXiv:2205.04559v1 [cs.CL])
cs.LG updates on arXiv.org arxiv.org
There has been significant debate in the NLP community about whether or not
attention weights can be used as an explanation - a mechanism for interpreting
how important each input token is for a particular prediction. The validity of
"attention as explanation" has so far been evaluated by computing the rank
correlation between attention-based explanations and existing feature
attribution explanations using LSTM-based models. In our work, we (i) compare
the rank correlation between five more recent feature attribution methods and …
artificial artificial intelligence arxiv evaluation explainable artificial intelligence intelligence language language processing natural natural language natural language processing processing