Towards the Reusability and Compositionality of Causal Representations
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2024
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866916160148078592 |
|---|---|
| author | Talon, Davide Lippe, Phillip James, Stuart Del Bue, Alessio Magliacane, Sara |
| author_facet | Talon, Davide Lippe, Phillip James, Stuart Del Bue, Alessio Magliacane, Sara |
| contents | 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. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2403_09830 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Towards the Reusability and Compositionality of Causal Representations Talon, Davide Lippe, Phillip James, Stuart Del Bue, Alessio Magliacane, Sara Machine Learning Artificial Intelligence 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. |
| title | Towards the Reusability and Compositionality of Causal Representations |
| topic | Machine Learning Artificial Intelligence |
| url | https://arxiv.org/abs/2403.09830 |