Towards Extracting Faithful and Descriptive Representations of Latent Variable Models

By Vicente Iván Sánchez Carmona, Tim Rocktaschel, Sebastian Riedel, Sameer Singh. 2015

40 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.

1. Sarah Tan
2. Giles Hooker
3. R. Caruana
4. Albert Gordo
5. Hao Peng
6. Qian Li
7. Cynthia Rudin
8. Paul Koch

Citations in the corpus

• Anchors: High-Precision Model-Agnostic Explanations (2018, background): “a decision tree (Craven and Shavlik 1996) or a set of rules (Sanchez et al. 2015)), but this may yield low human precision.”
• Model-Agnostic Interpretability of Machine Learning (2016, background): “ocally (but not globally) faithful. Discussion Some approaches are model agnostic by approximating the black box model by an interpretable one globally (Craven & Shavlik,1996;Baehrens et al.,2010;Sanchez et al., 2015). Global explanation, however, are often either not interpretable, or too simpl”
• Towards Faithfully Interpretable NLP Systems: How Should We Define and Evaluate Faithfulness? (2020, neutral): “, 2019) or “descriptive” (Carmona et al., 2015; Biecek, 2018) the interpretation is, or as “descriptive accuracy” (Murdoch et al.”
• Efficient Search for Diverse Coherent Explanations (2019, background): “in the introduction, but also raises the question of how accurate these approximations really are. These approximate models are either fitted globally (Craven and Shavlik, 1996; Martens et al., 2007; Sanchez et al., 2015) over the entire space of valid datapoints, or as a local approximation (Lundbe”
• A Comprehensive Survey on Automatic Knowledge Graph Construction (2023, background): “[280] train Bayesian networks with first-order logic to extract rules from embedding models.”
• Learning Global Additive Explanations for Neural Nets Using Model Distillation (2018, background): “, 2016; Frosst & Hinton, 2017) or model extraction (Fu, 1994; Sanchez et al., 2015; Lakkaraju et al., 2017; Bastani et al., 2017; Ribeiro et al., 2018).”
• Association for Computational Linguistics " Why Should I Trust You? " Explaining the Predictions of Any Classifier (2016, unknown): “Current work for explaining relies either on using interpretable models (such as decision trees [7] or additive models [5]) or on approximating black-box model predictions with interpretable models globally [2, 17].”
• Explanation Mining: Post Hoc Interpretability of Latent Factor Models for Recommendation Systems (2018, methodology): “We investigate these questions as we expand upon the work done in [33] by Research Track Paper KDD 2018, August 19‒23, 2018, London, United Kingdom”