Built independently by an author, for readers. Read the story and support ChapterPal
ChapterPal Logo
Log inSign up

keyword

CausalCircuit dataset

The CausalCircuit dataset is a synthetic benchmark dataset designed to evaluate causal representation learning and causal discovery methods from visual data. It comprises rendered images depicting a simulated robotic arm interacting with a causally connected system of buttons and lights. The underlying environment is governed by four ground-truth causal variables describing the position of the robotic arm along an arc and the intensities of red, green, and blue lights. Featuring paired image observations before and after random, unknown interventions, the dataset allows researchers to evaluate how effectively machine learning models can identify latent causal variables and recover the underlying causal structure directly from unstructured pixel data under weak supervision.

1 item

Weakly supervised causal representation learning

Weakly supervised causal representation learning

Johann Brehmer, Pim de Haan, Phillip Lippe, Taco S. Cohen

OrganizationsQualcommQUVA LabUniversity of Amsterdam

Why you should read this

Proves that high-level causal variables and mechanisms can be identified from pixel-level data paired across unknown interventions, and introduces implicit latent causal models to learn these structures without optimizing discrete graphs.

Learning high-level causal representations together with a causal model from unstructured low-level data such as pixels is impossible from observational data alone. We prove under mild assumptions that this representation is however identifiable in a weakly supervised setting. This involves a dataset with paired samples before and after random, unknown interventions, but no further labels. We then introduce implicit latent causal models, variational autoencoders that represent causal variables and causal structure without having to optimize an explicit discrete graph structure. On simple image data, including a novel dataset of simulated robotic manipulation, we demonstrate that such models can reliably identify the causal structure and disentangle causal variables.

Added

2026-09-30