Year: 2023
Type: conference-paper
Abstract: Recent model editing techniques promise to mitigate the problem of memorizing false or outdated associations during large language model (LLM) training.However, we show that these techniques can introduce large unwanted side effects which are not detected by existing specificity benchmarks.We extend the existing COUNTERFACT benchmark to include a dynamic component and dub our benchmark COUNTERFACT+.Additionally, we extend the metrics used for measuring specificity by a principled KL divergence-based metric.We use this improved benchmark to evaluate recent model editing techniques and find that they suffer from low specificity.Our findings highlight the need for improved specificity benchmarks that identify and prevent unwanted side effects.
Institutions: University of Edinburgh, University of Oxford
Language: English
Topic: Topic Modeling
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: hybrid
OpenAlex ID: W4385571289
