Compositional Attention Networks for Machine Reasoning

We present the MAC network, a novel fully differentiable neural network architecture, designed to facilitate explicit and expressive reasoning. MAC moves away from monolithic black-box neural architectures towards a design that encourages both transparency and versatility... (read more)

PDF Abstract ICLR 2018 PDF ICLR 2018 Abstract
TASK DATASET MODEL METRIC NAME METRIC VALUE GLOBAL RANK RESULT BENCHMARK
Visual Question Answering CLEVR MAC Accuracy 98.9 # 5
Visual Question Answering CLEVR-Humans MAC Accuracy 81.5 # 2

Methods used in the Paper


METHOD TYPE
Interpretability
Image Models