A Fresh Look at Sanity Checks for Saliency Maps

arXiv:2405.02383v1 Announce Type: cross
Abstract: The Model Parameter Randomisation Test (MPRT) is highly recognised in the eXplainable Artificial Intelligence (XAI) community due to its fundamental evaluative criterion: explanations should be sensitive to the parameters of the model they seek to explain. However, recent studies have raised several methodological concerns for the empirical interpretation of MPRT. In response, we propose two modifications to the original test: Smooth MPRT and Efficient MPRT. The former reduces the impact of noise on evaluation outcomes …

abstract artificial artificial intelligence arxiv checks community concerns criterion cs.ai cs.cv cs.lg explainable artificial intelligence fundamental however intelligence look maps parameters seek stat.ml studies test type xai