Deep learning models are ‘fragile’, BoE paper finds

Models suffering from “Rashomon effect” need better supervision, authors say

AI analytics

Deep learning models are “fragile” and could create serious risks at financial firms that use them, a working paper published by the Bank of England finds.

The team of researchers – from the BoE, Reserve Bank of Australia and University College London – study not only the predictions of deep learning models but also explanations of how each model arrives at particular results. Deep learning is a form of artificial intelligence that uses models structured into multiple layers.

The authors find

Only users who have a paid subscription or are part of a corporate subscription are able to print or copy content.

To access these options, along with all other subscription benefits, please contact or view our subscription options here:

You are currently unable to copy this content. Please contact to find out more.

Sorry, our subscription options are not loading right now

Please try again later. Get in touch with our customer services team if this issue persists.

New to Central Banking? View our subscription options

Register for Central Banking

All fields are mandatory unless otherwise highlighted

This address will be used to create your account

You need to sign in to use this feature. If you don’t have a Central Banking account, please register for a trial.

Sign in
You are currently on corporate access.

To use this feature you will need an individual account. If you have one already please sign in.

Sign in.

Alternatively you can request an individual account