Pre-trained models, data, code & materials from the paper "ImageNet-trained CNNs are biased towards texture; increasing shape bias improves accuracy and robustness" (ICLR 2019 Oral)

Data, code & materials from the paper "Generalisation in humans and deep neural networks" (NeurIPS 2018)

Error consistency: a black-box analysis for comparing errors between decision makers (NeurIPS 2020)