Scattering Compositional Learner

PyTorch implementation of Scattering Compositional Learner [1] for solving Raven's Progressive Matrices.

Setup

$ pip install scattering_compositional_learner

Usage

import torch

from scattering_compositional_learner import ScatteringCompositionalLearner

x = torch.rand(4, 16, 160, 160)
scl = ScatteringCompositionalLearner(image_size=160)
logits = scl(x)
y_hat = logits.log_softmax(dim=-1)
y_hat  # torch.Tensor with shape (4, 8)

Unit tests

$ python -m pytest tests

Alternative implementations

The same model was additionally implemented by:

• paper authors
• lucidrains

Bibliography

[1] Wu, Yuhuai, et al. "The Scattering Compositional Learner: Discovering Objects, Attributes, Relationships in Analogical Reasoning." arXiv preprint arXiv:2007.04212 (2020).

Citations

@article{wu2020scattering,
  title={The Scattering Compositional Learner: Discovering Objects, Attributes, Relationships in Analogical Reasoning},
  author={Wu, Yuhuai and Dong, Honghua and Grosse, Roger and Ba, Jimmy},
  journal={arXiv preprint arXiv:2007.04212},
  year={2020}
}