Contrastive Explanation

By P. Lipton. 1990

Subfields: Safety, Robustness & Interpretability, Philosophy

How researchers see this paper

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

In Safety, Robustness & Interpretability

In Philosophy

Citations in the corpus

Explanation in Human-AI Systems: A Literature Meta-Review, Synopsis of Key Ideas and Publications, and Bibliography for Explainable AI (2019, background): “…Byrne (1997; 2017); Byrne and McEleney (2000); Kahneman and Varey (1990); Halpern and Pearl (2005a,b), Hilton, McClure and Slugoski (2005); Hitchcock (2003); Hitchcock and Woodward (2003); Hoffman, et al., (2014); Khemlani and Johnson-Laird (2010); Kim, Khanna, and Koyejo (2016); Lipton,…”
Explainable AI: Beware of Inmates Running the Asylum Or: How I Learnt to Stop Worrying and Love the Social and Behavioural Sciences (2017, background): “However, it is also an opportunity: as Lipton [1990] argues, answering a contrastive question is often easier than giving a full cause attribution because one only needs to understand the diﬀerence between the two cases, so one can provide a complete explanation without determining or even knowing…”
Explanation in Artificial Intelligence: Insights from the Social Sciences (2017, background): “We take the definition that an explanation is an answer to a why–question [35, 138, 99, 102].”
Explaining Explanations in AI (2018, neutral): “However, very few stop to check what they are agreeing to, and to find out what explainable AI means to other people involved in the discussion (Lipton, 2016).”
Aligning Faithful Interpretations with their Social Attribution (2020, methodology): “We refer to P as the fact and Q as the foil (Lipton, 1990).”
Prompting Contrastive Explanations for Commonsense Reasoning Tasks (2021, background): “It explains why some fact P occurred instead of some foil Q, where Q can be implicit (Hesslow, 1988; Lipton, 1990; Miller, 2019).”
Explaining NLP Models via Minimal Contrastive Editing (MiCE) (2020, background): “3 A long line of research in the cognitive sciences and philosophy has found that human explanations are contrastive (Van Fraassen, 1980; Lipton, 1990; Miller, 2019).”
Generating Counterfactual and Contrastive Explanations using SHAP (2019, background): “…of artiﬁcial intelligence, the former is asking why a particular algorithm gave an output rather than some other output that the questioner expected, while the latter is asking why an algorithm gave a particular output this time but some (probably different) output another time [Lipton, 1990].”
Scientific Explanation And The Sense Of Understanding* (2002, background): “In contemporary science, global unification arises spontaneously from coordinated piecemeal efforts, not from a meta-level at which the philosopher or reflective scientist assembles remote domains (Miller, 1987).”
Problems with Shapley-value-based explanations as feature importance measures (2020, background): “He attributes this insight to work by Lipton (1990).”
Designing Theory-Driven User-Centric Explainable AI (2019, background): “A user seeking an explanation may know that the system output is not as expected and therefore seek a contrastive explanation [70, 78]. and allowing details on demand, but mainly to limit information overload (e.g., [27, 57, 68]).”
A Survey of Contrastive and Counterfactual Explanation Generation Methods for Explainable Artificial Intelligence (2021, methodology): “[61] refer to Lipton’s definition of contrastive explanation [12].”
Combinatorial Methods for Explainable AI (2020, background): “This is suggested as a potential approach to explainable AI, because providing a contrast case may be easier than a full set of causes [10][15].”
Evaluating XAI: A comparison of rule-based and example-based explanations (2021, background): “Human explanations tend to be contrastive: they compare a certain phenomenon (fact) with a hypothetical one (foil) [33,34].”