Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
119 tokens/sec
GPT-4o
56 tokens/sec
Gemini 2.5 Pro Pro
43 tokens/sec
o3 Pro
6 tokens/sec
GPT-4.1 Pro
47 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

Delivering Inflated Explanations (2306.15272v1)

Published 27 Jun 2023 in cs.AI and cs.LG

Abstract: In the quest for Explainable Artificial Intelligence (XAI) one of the questions that frequently arises given a decision made by an AI system is, ``why was the decision made in this way?'' Formal approaches to explainability build a formal model of the AI system and use this to reason about the properties of the system. Given a set of feature values for an instance to be explained, and a resulting decision, a formal abductive explanation is a set of features, such that if they take the given value will always lead to the same decision. This explanation is useful, it shows that only some features were used in making the final decision. But it is narrow, it only shows that if the selected features take their given values the decision is unchanged. It's possible that some features may change values and still lead to the same decision. In this paper we formally define inflated explanations which is a set of features, and for each feature of set of values (always including the value of the instance being explained), such that the decision will remain unchanged. Inflated explanations are more informative than abductive explanations since e.g they allow us to see if the exact value of a feature is important, or it could be any nearby value. Overall they allow us to better understand the role of each feature in the decision. We show that we can compute inflated explanations for not that much greater cost than abductive explanations, and that we can extend duality results for abductive explanations also to inflated explanations.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (50)
  1. Axiomatic foundations of explainability. In IJCAI, pages 636–642, 2022.
  2. Leila Amgoud. Non-monotonic explanation functions. In ECSQARU, pages 19–31, 2021.
  3. Leila Amgoud. Explaining black-box classifiers: Properties and functions. Int. J. Approx. Reason., 155:40–65, 2023.
  4. On computing probabilistic explanations for decision trees. In NeurIPS, 2022.
  5. On the glucose SAT solver. Int. J. Artif. Intell. Tools, 27(1):1840001:1–1840001:25, 2018.
  6. On tractable XAI queries based on compiled representations. In KR, pages 838–849, 2020.
  7. On the computational intelligibility of boolean classifiers. In KR, pages 74–86, 2021.
  8. On the explanatory power of boolean decision trees. Data Knowl. Eng., 142:102088, 2022.
  9. On pixel-wise explanations for non-linear classifier decisions by layer-wise relevance propagation. PloS one, 10(7):e0130140, 2015.
  10. Towards formal XAI: formally approximate minimal explanations of neural networks. In TACAS, pages 187–207, 2023.
  11. An almost optimal algorithm for unbounded searching. Inf. Process. Lett., 5(3):82–87, 1976.
  12. A symbolic approach for counterfactual explanations. In SUM, pages 270–277, 2020.
  13. ASTERYX: A model-agnostic sat-based approach for symbolic and score-based explanations. In CIKM, pages 120–129, 2021.
  14. On symbolically encoding the behavior of random forests. CoRR, abs/2007.01493, 2020.
  15. On the tractability of explaining decisions of classifiers. In CP, pages 21:1–21:18, 2021.
  16. Tractability of explaining classifier decisions. Artif. Intell., 316:103841, 2023.
  17. On the (complete) reasons behind decisions. J. Log. Lang. Inf., 32(1):63–88, 2023.
  18. Minimum satisfying assignments for SMT. In CAV, pages 394–409, 2012.
  19. Sufficient reasons for classifier decisions in the presence of domain constraints. In AAAI, February 2022.
  20. On efficiently explaining graph-based classifiers. In KR, pages 356–367, 2021.
  21. SAT-based rigorous explanations for decision lists. In SAT, pages 251–269, 2021.
  22. PySAT: A python toolkit for prototyping with SAT oracles. In SAT, pages 428–437, 2018.
  23. Abduction-based explanations for machine learning models. In AAAI, pages 1511–1519, 2019.
  24. From contrastive to abductive explanations and back again. In AIxIA, pages 335–355, 2020.
  25. Alexey Ignatiev. Towards trustable explainable AI. In IJCAI, pages 5154–5158, 2020.
  26. On explaining random forests with SAT. In IJCAI, pages 2584–2591, 2021.
  27. On tackling explanation redundancy in decision trees. J. Artif. Intell. Res., 75:261–321, 2022.
  28. A new class of explanations for classifiers with non-binary features. CoRR, abs/2304.14760, 2023.
  29. A logic of ”black box” classifier systems. In WoLLIC, pages 158–174, 2022.
  30. A unified logical framework for explanations in classifier systems. J. Log. Comput., 33(2):485–515, 2023.
  31. A unified approach to interpreting model predictions. In NeurIPS, pages 4765–4774, 2017.
  32. On guaranteed optimal robust explanations for NLP models. In IJCAI, pages 2658–2665, 2021.
  33. Delivering trustworthy AI through formal XAI. In AAAI, pages 12342–12350, 2022.
  34. Explanations for monotonic classifiers. In ICML, pages 7469–7479, 2021.
  35. João Marques-Silva. Logic-based explainability in machine learning. CoRR, abs/2211.00541, 2022.
  36. Tim Miller. Explanation in artificial intelligence: Insights from the social sciences. Artif. Intell., 267:1–38, 2019.
  37. Christoph Molnar. Interpretable Machine Learning. Leanpub, 2020. http://tiny.cc/6c76tz.
  38. PMLB: a large benchmark suite for machine learning evaluation and comparison. BioData Mining, 10(1):36, 2017.
  39. Fabian Pedregosa and et al. Scikit-learn: Machine learning in Python. Journal of Machine Learning Research, 12:2825–2830, 2011.
  40. ”why should I trust you?”: Explaining the predictions of any classifier. In KDD, pages 1135–1144, 2016.
  41. Anchors: High-precision model-agnostic explanations. In AAAI, pages 1527–1535, 2018.
  42. Interpretable machine learning: Fundamental principles and 10 grand challenges. Statistics Surveys, 16:1–85, 2022.
  43. Cynthia Rudin. Stop explaining black box machine learning models for high stakes decisions and use interpretable models instead. Nature Machine Intelligence, 1(5):206–215, 2019.
  44. Towards explainable artificial intelligence. In Samek et al. Samek et al. (2019), pages 5–22.
  45. Explainable AI: Interpreting, Explaining and Visualizing Deep Learning. Springer, 2019.
  46. Explaining deep neural networks and beyond: A review of methods and applications. Proc. IEEE, 109(3):247–278, 2021.
  47. Toward verified artificial intelligence. Commun. ACM, 65(7):46–55, 2022.
  48. A symbolic approach to explaining bayesian network classifiers. In IJCAI, pages 5103–5111, 2018.
  49. UCI Machine Learning Repository. https://archive.ics.uci.edu/ml, 2020.
  50. The computational complexity of understanding binary classifier decisions. J. Artif. Intell. Res., 70:351–387, 2021.
User Edit Pencil Streamline Icon: https://streamlinehq.com
Authors (4)
  1. Yacine Izza (16 papers)
  2. Alexey Ignatiev (29 papers)
  3. Peter Stuckey (8 papers)
  4. Joao Marques-Silva (67 papers)
Citations (4)
View on Semantic Scholar

Summary

We haven't generated a summary for this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Summarize Now