[2509. 22072] Fine-tuning Done Right in Model Editing - arXiv. org By clarifying a long-standing misconception and introducing a principled localized tuning strategy, we advance fine-tuning from an underestimated baseline to a leading method for model editing, establishing a solid foundation for future research
Fine-tuning Done Right in Model Editing | OpenReview TL;DR: This paper corrects prevailing misconceptions about the effectiveness of fine-tuning in model editing, and demonstrates that simple fine-tuning can effectively address model editing tasks
Fine-tuning Done Right in Model Editing - Semantic Scholar By clarifying a long-standing misconception and introducing a principled localized tuning strategy, we advance fine-tuning from an underestimated baseline to a leading method for model editing, establishing a solid foundation for future research
Paper page - Fine-tuning Done Right in Model Editing This paper revives fine-tuning for model editing, proving that by simply fixing its flawed implementation and applying a localized parameter update strategy, the simplest method becomes state-of-the-art, handling an unprecedented 100K sequential edits on models up to 72B parameters