Physical Reasoning Using Dynamics Aware Embeddings

In International Conference on Machine Learning Workshop on Self-Supervised Learning for Reasoning and Perception, 2021

Eltayeb Ahmed, Anton Bakhtin, Laurens van der Maaten, Rohit Girdhar

Teaser figure for Physical Reasoning Using Dynamics Aware Embeddings

Abstract

A common approach to solving physicalreasoning tasks is to train a value learner on example tasks. A limitation of such an approach is that it requires learning about object dynamics solely from reward values assigned to the final state of a rollout of the environment. This study aims to address this limitation by augmenting the reward value with self-supervised signals about object dynamics. Specifically, we train the model to characterize the similarity of two environment rollouts, jointly with predicting the outcome of the reasoning task. This similarity can be defined as a distance measure between the trajectory of objects in the two rollouts, or learned directly from pixels using a contrastive formulation. Empirically, we find that this approach leads to substantial performance improvements on the PHYRE benchmark for physical reasoning, establishing a new state-of-the-art.

Citation

@article{ahmed2021physical,
  title={Physical Reasoning Using Dynamics-Aware Models},
  author={Ahmed, Eltayeb and Bakhtin, Anton and van der Maaten, Laurens and Girdhar, Rohit},
  booktitle={ICML Workshop},
  year={2021}
}