Using a Cross-Task Grid of Linear Probes to Interpret CNN Model Predictions On Retinal Images

Kavli Affiliate: Michael P. Brenner

| First 5 Authors: Katy Blumer, Subhashini Venugopalan, Michael P. Brenner, Jon Kleinberg,

| Summary:

We analyze a dataset of retinal images using linear probes: linear regression
models trained on some "target" task, using embeddings from a deep
convolutional (CNN) model trained on some "source" task as input. We use this
method across all possible pairings of 93 tasks in the UK Biobank dataset of
retinal images, leading to ~164k different models. We analyze the performance
of these linear probes by source and target task and by layer depth. We observe
that representations from the middle layers of the network are more
generalizable. We find that some target tasks are easily predicted irrespective
of the source task, and that some other target tasks are more accurately
predicted from correlated source tasks than from embeddings trained on the same

| Search Query: ArXiv Query: search_query=au:”Michael P. Brenner”&id_list=&start=0&max_results=10

Read More

Leave a Reply