1 paper · 1 filter
Guilin Zhang, Kai Zhao, Jeffrey Friedman +3
Robust explanations are increasingly required for user trust in enterprise NLP, yet pre-deployment validation is difficult in the common case of black-box deployment (API-only acce…