TY - GEN
T1 - Are disentangled representations helpful for abstract visual reasoning?
AU - van Steenkiste, Sjoerd
AU - Locatello, Francesco
AU - Schmidhuber, Jürgen
AU - Bachem, Olivier
N1 - Generated from Scopus record by KAUST IRTS on 2022-09-14
PY - 2019/1/1
Y1 - 2019/1/1
N2 - A disentangled representation encodes information about the salient factors of variation in the data independently. Although it is often argued that this representational format is useful in learning to solve many real-world down-stream tasks, there is little empirical evidence that supports this claim. In this paper, we conduct a large-scale study that investigates whether disentangled representations are more suitable for abstract reasoning tasks. Using two new tasks similar to Raven's Progressive Matrices, we evaluate the usefulness of the representations learned by 360 state-of-the-art unsupervised disentanglement models. Based on these representations, we train 3600 abstract reasoning models and observe that disentangled representations do in fact lead to better down-stream performance. In particular, they enable quicker learning using fewer samples.
AB - A disentangled representation encodes information about the salient factors of variation in the data independently. Although it is often argued that this representational format is useful in learning to solve many real-world down-stream tasks, there is little empirical evidence that supports this claim. In this paper, we conduct a large-scale study that investigates whether disentangled representations are more suitable for abstract reasoning tasks. Using two new tasks similar to Raven's Progressive Matrices, we evaluate the usefulness of the representations learned by 360 state-of-the-art unsupervised disentanglement models. Based on these representations, we train 3600 abstract reasoning models and observe that disentangled representations do in fact lead to better down-stream performance. In particular, they enable quicker learning using fewer samples.
UR - http://www.scopus.com/inward/record.url?scp=85089890854&partnerID=8YFLogxK
M3 - Conference contribution
BT - Advances in Neural Information Processing Systems
PB - Neural information processing systems foundation
ER -