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 current trend toward ever-larger models makes standard retrainingprocedures an ever-more expensive burden. For this reason, there is growinginterest in model editing, which enables computationally inexpensive,interpretable, post-hoc model modifications. While many model editingtechniques are promising, research on the properties of edited models islargely limited to evaluation of validation accuracy. The robustness of editedmodels is an important and yet mostly unexplored topic. In this paper, weemploy recently developed techniques from the field of deep learning robustnessto investigate both how model editing affects the general robustness of amodel, as well as the robustness of the specific behavior targeted by the edit.We find that edits tend to reduce general robustness, but that the degree ofdegradation depends on the editing algorithm and layers chosen. Motivated bythese observations we introduce a new model editing algorithm, 1-layerinterpolation (1-LI), which uses weight-space interpolation to navigate thetrade-off between editing task accuracy and general robustness.

Edit at your own risk: evaluating the robustness of edited models to distribution shifts