arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2303.00046cs.LG
收藏

编辑需自担风险:评估被编辑模型对分布偏移的鲁棒性

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

Pacific Northwest National Laboratory(太平洋西北国家实验室);University of Washington(华盛顿大学)

Davis Brown, Charles Godfrey, Cody Nizinski, Jonathan Tu, Henry Kvinge

AI总结:

本文研究模型编辑在分布偏移下的鲁棒性,发现编辑会降低总体鲁棒性且退化程度因算法和层而异,并提出利用权重空间插值平衡准确率与鲁棒性的 1-LI 算法。

摘要:

当前模型规模不断增大的趋势,使标准重训练流程成为越来越昂贵的负担。因此,人们对模型编辑的兴趣日益增长,因为它能够实现计算成本低、可解释的事后模型修改。尽管许多模型编辑技术很有前景,但关于被编辑模型性质的研究在很大程度上仍局限于验证准确率评估。被编辑模型的鲁棒性是一个重要却基本未被探索的课题。本文采用深度学习鲁棒性领域近期发展出的技术,考察模型编辑如何影响模型的总体鲁棒性,以及编辑所针对的特定行为的鲁棒性。我们发现,编辑往往会降低总体鲁棒性,但退化程度取决于所选择的编辑算法和层。受这些观察启发,我们提出一种新的模型编辑算法 1-layer interpolation(1-LI),该算法利用权重空间插值来协调编辑任务准确率与总体鲁棒性之间的权衡。

英文摘要:

The current trend toward ever-larger models makes standard retraining procedures an ever-more expensive burden. For this reason, there is growing interest in model editing, which enables computationally inexpensive, interpretable, post-hoc model modifications. While many model editing techniques are promising, research on the properties of edited models is largely limited to evaluation of validation accuracy. The robustness of edited models is an important and yet mostly unexplored topic. In this paper, we employ recently developed techniques from the field of deep learning robustness to investigate both how model editing affects the general robustness of a model, 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 of degradation depends on the editing algorithm and layers chosen. Motivated by these observations we introduce a new model editing algorithm, 1-layer interpolation (1-LI), which uses weight-space interpolation to navigate the trade-off between editing task accuracy and general robustness.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑