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

The generalization of machine learning (ML) models to out-of-distribution(OOD) examples remains a key challenge in extracting information from upcomingastronomical surveys. Interpretability approaches are a natural way to gaininsights into the OOD generalization problem. We use Centered Kernel Alignment(CKA), a similarity measure metric of neural network representations, toexamine the relationship between representation similarity and performance ofpre-trained Convolutional Neural Networks (CNNs) on the CAMELS MultifieldDataset. We find that when models are robust to a distribution shift, theyproduce substantially different representations across their layers on OODdata. However, when they fail to generalize, these representations change lessfrom layer to layer on OOD data. We discuss the potential application ofsimilarity representation in guiding model design, training strategy, andmitigating the OOD problem by incorporating CKA as an inductive bias duringtraining.

Towards out-of-distribution generalization in large-scale astronomical surveys: robust networks learn similar representations