Minimal Model Explanations

By Robert W. Batterman, Collin Rice. Philosophia Scientiæ, 2014

311 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Reasons, Values, Stakeholders: A Philosophical Framework for Explainable Artificial Intelligence (2021, background): “…some philosophers have argued that mathematical facts explain a variety of empirical phenomenon better than causal information does, across a range of sciences, from condensed matter physics to evolutionary and developmental biology or computational neuroscience [54, 55, 56, 61, 58, 57, 60].”
Why Attention is Not Explanation: Surgical Intervention and Causal Reasoning about Neural Models (2020, background): “Explanations are built by connecting models to phenomena via a “justiﬁcatory step,” whose details will be particular to the case at hand.”
Evaluating and Aggregating Feature-based Model Explanations (2020, background): “Since humans cannot process a lot of information at once, some have argued for minimal model explanations that contain only relevant and representative features [ Batterman and Rice, 2014 ] ; therefore, an explanations should not be complex (i.e., use few features).”