English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Representation learning assumes that real-world data is generated by a fewsemantically meaningful generative factors (i.e., sources of variation) andaims to discover them in the latent space. These factors are expected to becausally disentangled, meaning that distinct factors are encoded into separatelatent variables, and changes in one factor will not affect the values of theothers. Compared to statistical independence, causal disentanglement allowsmore controllable data generation, improved robustness, and bettergeneralization. However, most existing work assumes unconfoundedness in thediscovery process, that there are no common causes to the generative factorsand thus obtain only statistical independence. In this paper, we recognize theimportance of modeling confounders in discovering causal generative factors.Unfortunately, such factors are not identifiable without proper inductive bias.We fill the gap by introducing a framework entitled Confounded-Disentanglement(C-Disentanglement), the first framework that explicitly introduces theinductive bias of confounder via labels from domain expertise. In addition, weaccordingly propose an approach to sufficiently identify the causallydisentangled factors under any inductive bias of the confounder. We conductextensive experiments on both synthetic and real-world datasets. Our methoddemonstrates competitive results compared to various SOTA baselines inobtaining causally disentangled features and downstream tasks under domainshifts.

C-Disentanglement: Discovering Causally-Independent Generative Factors under an Inductive Bias of Confounder