Do explanations make VQA models more predictable to a human?

Arjun Chandrasekaran, Viraj Prabhu, Deshraj Yadav, Prithvijit Chattopadhyay, Devi Parikh

Browse the full EMNLP paper archive.