mlsanityData to 7 Oct 2026

TT-FSI: Scalable Faithful Shapley Interactions via Tensor-Train

arXiv 2601.01903 ↗First posted 5 Jan 2026cs.LGcited by 0

Not on the map. Why?

Abstract (arXiv)

The Faithful Shapley Interaction (FSI) index uniquely satisfies the faithfulness axiom among Shapley interaction indices, but computing FSI requires O(dℓ · 2d) time and existing implementations use O(4d) memory. We present TT-FSI, which exploits FSI's algebraic structure via Matrix Product Operators (MPO). Our main theoretical contribution is proving that the linear operator v mapsto FSI(v) admits an MPO representation with TT-rank O(ℓ d), enabling an efficient sweep algorithm with O(ℓ2 d3 · 2d) time and O(ℓ d2) core storage an exponential improvement over existing methods. Experiments on six datasets (d=8 to d=20) demonstrate up to 280× speedup over baseline, 85× over SHAP-IQ, and 290× memory reduction. TT-FSI scales to d=20 (1M coalitions) where all competing methods fail.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Builds on 21 of its 43 references that are in the map. First posted on arXiv 5 Jan 2026.

Builds on

21 of 43 references.

References · 21

  1. 2017A Unified Approach to Interpreting Model Predictionscited by 6,488
  2. 2016"Why Should I Trust You?": Explaining the Predictions of Any Classifiercited by 5,040
  3. 2017Axiomatic Attribution for Deep Networkscited by 3,350
  4. 2019Explainable Artificial Intelligence (XAI): Concepts, Taxonomies, Opportunities and Challenges toward Responsible AIcited by 1,543
  5. 2018A Survey Of Methods For Explaining Black Box Modelscited by 970
  6. 2019Fooling LIME and SHAP: Adversarial Attacks on Post hoc Explanation Methodscited by 440
  7. 2015Tensorizing Neural Networkscited by 246
  8. 2020Explaining by Removing: A Unified Framework for Model Explanationcited by 218
  9. 2019Explaining individual predictions when features are dependent: More accurate approximations to Shapley valuescited by 205
  10. 2019Feature relevance quantification in explainable AI: A causal problemcited by 164
  11. 2019The Shapley Taylor Interaction Indexcited by 143
  12. 2020True to the Model or True to the Data?cited by 123
9 more
  1. 2021FastSHAP: Real-Time Shapley Value Estimationcited by 123
  2. 2021Sampling Permutations for Shapley Value Estimationcited by 101
  3. 2020On the Tractability of SHAP Explanationscited by 92
  4. 2022Faith-Shap: The Faithful Shapley Interaction Indexcited by 68
  5. 2024shapiq: Shapley Interactions for Machine Learningcited by 42
  6. 2021On the Complexity of SHAP-Score-Based Explanations: Tractability via Knowledge Compilation and Non-Approximability Resultscited by 32
  7. 2024SVARM-IQ: Efficient Approximation of Any-order Shapley Interactions through Stratificationcited by 16
  8. 2017Duality of Graphical Models and Tensor Networkscited by 15
  9. 2025Interaction Tensor SHAPcited by 1

Built on by

Not cited yet.