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 2010.13233cs.LG
收藏

Now You See Me(CME):基于概念的模型抽取

Now You See Me (CME): Concept-based Model Extraction

The University of Cambridge(剑桥大学);The Alan Turing Institute(艾伦·图灵研究所)

Dmitry Kazhdan, Botty Dimanov, Mateja Jamnik, Pietro Liò, Adrian Weller

摘要:

深度神经网络(DNN)在一系列任务上取得了卓越的性能。进一步增强基于DNN的方法的关键一步是提高其可解释性。在这项工作中,我们提出了CME:一个基于概念的模型提取框架,用于通过基于概念的提取模型分析DNN模型。使用两个案例研究(dSprites和加州理工学院UCSD Birds),我们演示了CME如何用于(i)分析DNN模型学习的概念信息(ii)分析DNN在预测输出标签时如何使用此概念信息(iii)识别可以进一步提高DNN预测性能的关键概念信息(对于其中一个案例研究,我们展示了如何使用30%的可用概念将模型准确性提高14%以上)。

英文摘要:

Deep Neural Networks (DNNs) have achieved remarkable performance on a range of tasks. A key step to further empowering DNN-based approaches is improving their explainability. In this work we present CME: a concept-based model extraction framework, used for analysing DNN models via concept-based extracted models. Using two case studies (dSprites, and Caltech UCSD Birds), we demonstrate how CME can be used to (i) analyse the concept information learned by a DNN model (ii) analyse how a DNN uses this concept information when predicting output labels (iii) identify key concept information that can further improve DNN predictive performance (for one of the case studies, we showed how model accuracy can be improved by over 14%, using only 30% of the available concepts).

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。

↑