Year: 2025
Type: conference-paper
Abstract: Model editing has been gaining increasing attention over the past few years.For Knowledge Editing in particular, more challenging evaluation datasets have recently been released.These datasets use different methodologies to score the success of editors.Yet, it remains underexplored how robust these methodologies are and whether they unfairly favor some editors.Moreover, the disruptive impact of these editors on overall model capabilities remains a constant blind spot.We address both of these problems and show that choosing different metrics and evaluation methodologies as well as different edit batch sizes can lead to a different ranking of knowledge editors.Crucially we demonstrate this effect also on general language understanding tasks evaluated alongside the knowledge editing tasks.Further we include a manual assessment of the string matching based evaluation method for knowledge editing that is favored by recently released datasets, revealing a tendency to produce false positive matches.
Authors: Sebastian Pohl, Max Ploner, Alan Akbik
Institution: Humboldt-Universität zu Berlin
Language: English
Topic: Topic Modeling
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
SDG: Quality education
Open Access status: gold
Funder: Deutsche Forschungsgemeinschaft
OpenAlex ID: W4412944286