Towards the Reusability and Compositionality of Causal Representations
This work addresses the challenge of reusability and compositionality in causal representation learning for AI systems, representing an incremental advance by extending existing methods to multi-environment settings.
The paper tackles the problem of adapting causal representations learned from temporal image sequences to new environments, proposing DECAF to identify reusable and adaptable causal factors, and shows it enables accurate representations with few samples in experiments on three benchmarks.
Causal Representation Learning (CRL) aims at identifying high-level causal factors and their relationships from high-dimensional observations, e.g., images. While most CRL works focus on learning causal representations in a single environment, in this work we instead propose a first step towards learning causal representations from temporal sequences of images that can be adapted in a new environment, or composed across multiple related environments. In particular, we introduce DECAF, a framework that detects which causal factors can be reused and which need to be adapted from previously learned causal representations. Our approach is based on the availability of intervention targets, that indicate which variables are perturbed at each time step. Experiments on three benchmark datasets show that integrating our framework with four state-of-the-art CRL approaches leads to accurate representations in a new environment with only a few samples.