Democratizing Evaluation of Deep Model Interpretability through Consensus

1 Jan 2021  ·  Xuhong LI, Haoyi Xiong, Siyu Huang, Shilei Ji, Yanjie Fu, Dejing Dou ·

Deep learning interpretability tools, such as (Bau et al., 2017; Ribeiro et al., 2016; Smilkov et al., 2017), have been proposed to explain and visualize the ways that deep neural networks make predictions. The success of these methods highly relies on human subjective interpretations, i.e., the ground truth of interpretations, such as feature importance ranking or locations of visual objects, when evaluating the interpretability of the deep models on a specific task. For tasks that the ground truth of interpretations is not available, we propose a novel framework Consensus incorporating an ensemble of deep models as the committee for interpretability evaluation. Given any task/dataset, Consensus first obtains the interpretation results using existing tools, e.g., LIME (Ribeiro et al., 2016), for every model in the committee, then aggregates the results from the entire committee and approximates the “ground truth” of interpretations through voting. With such approximated ground truth, Consensus evaluates the interpretability of a model through matching its interpretation result and the approximated one, and ranks the matching scores together with committee members, so as to pursue the absolute and relative interpretability evaluation results. We carry out extensive experiments to validate Consensus on various datasets. The results show that Consensus can precisely identify the interpretability for a wide range of models on ubiquitous datasets that the ground truth is not available. Robustness analyses further demonstrate the advantage of the proposed framework to reach the consensus of interpretations through simple voting and evaluate the interpretability of deep models. Through the proposed Consensus framework, the interpretability evaluation has been democratized without the need of ground truth as criterion.

PDF Abstract
No code implementations yet. Submit your code now

Results from the Paper


  Submit results from this paper to get state-of-the-art GitHub badges and help the community compare results to other papers.

Methods