Efficiently Editing Mixture-of-Experts Models with Compressed Experts

Open in new window