Sanity Checking Causal Representation Learning on a Simple Real-World System

TL;DR

This paper evaluates causal representation learning methods on a controlled real-world optical system with known causal factors, revealing significant failures and reproducibility issues that challenge their practical applicability.

Contribution

It provides a benchmark and analysis highlighting the limitations of current CRL methods in real-world settings, emphasizing the need for more robust approaches.

Findings

All evaluated methods failed to recover causal factors.

Most methods failed even on simplified synthetic data.

Assumptions on mixing functions are crucial but often violated.

Abstract

We evaluate methods for causal representation learning (CRL) on a simple, real-world system where these methods are expected to work. The system consists of a controlled optical experiment specifically built for this purpose, which satisfies the core assumptions of CRL and where the underlying causal factors (the inputs to the experiment) are known, providing a ground truth. We select methods representative of different approaches to CRL and find that they all fail to recover the underlying causal factors. To understand the failure modes of the evaluated algorithms, we perform an ablation on the data by substituting the real data-generating process with a simpler synthetic equivalent. The results reveal a reproducibility problem, as most methods already fail on this synthetic ablation despite its simple data-generating process. Additionally, we observe that common assumptions on the mixing function are crucial for the performance of some of the methods but do not hold in the real data. Our efforts highlight the contrast between the theoretical promise of the state of the art and the challenges in its application. We hope the benchmark serves as a simple, real-world sanity check to further develop and validate methodology, bridging the gap towards CRL methods that work in practice. We make all code and datasets publicly available at github.com/simonbing/CRLSanityCheck