SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2204.00762 · CVPR · 2022

Do learned representations respect causal relationships?

Lan Wang, Vishnu Boddeti

arXiv · PDF · Open in the Atlas

Code that ran

We have not lifted any functions out of this paper's repositories yet, so there is nothing we have run. The repositories linked to it are listed below.

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

Data often has many semantic attributes that are causally associated with each other. But do attributespecific learned representations of data also respect the same causal relations? We answer this question in three steps. First, we introduce NCINet, an approach for observational causal discovery from high-dimensional data. It is trained purely on synthetically generated representations and can be applied to real representations, and is specifically designed to mitigate the domain gap between the two. Second, we apply NCINet to identify the causal relations between image representations of different pairs of attributes with known and unknown causal relations between the labels. For this purpose, we consider image representations learned for predicting attributes on the 3D Shapes, CelebA, and the CASIA-WebFace datasets, which we annotate with multiple multi-class attributes. Third, we analyze the effect on the underlying causal relation between learned representations induced by various design choices in representation learning. Our experiments indicate that (1) NCINet significantly outperforms existing observational causal discovery approaches for estimating the causal relation between pairs of random samples, both in the presence and absence of an unobserved confounder, (2) under controlled scenarios, learned representations can indeed satisfy the underlying causal relations between their respective labels, and (3) the causal relations are positively correlated with the predictive capability of the representations. Code and annotations are available at: https://github. com / human -analysis / causal -relationsbetween-representations.

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2204.00762")
get_code_for_paper("2204.00762")
have("2204.00762")

The run record, dated, one paper per request, free:

curl https://syntology.ai/api/ran/2204.00762.json

A badge for a README (the split and the date, never a ratio):

[![Syntology run record](https://syntology.ai/api/ran/2204.00762.svg)](https://syntology.ai/paper/2204.00762)

Connect an agent — have() is free.