AI Paper Atlas

Research map / Graph learning / Graph algorithms and clustering

TN-SHAP-G: Graph-Structured Tensor Network Surrogates for Shapley Values and Interactions

Farzaneh Heidari, Guillaume Rabusseau

ICML 2026 · Graph learning › Graph algorithms and clustering · approach: Continuous DAG structure learning · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
use structure-aware coalition sampling (edge-flip and triangle-flip neighbors) instead of uniform (ablation)TN-SHAP-G · higher-order (O3) interaction recoveryO3 cosine similarity: 0.48 -> 0.91Improved
use graph-aligned tensor network instead of tensor train (TT) under parameter matching (ablation)TN-SHAP-G · molecular graphs, low query budget (10n coalitions)O1/O2 cosine similarityImproved
vary TT node ordering (default, BFS, random) in the tensor train baseline (ablation)TT surrogate · molecular graphsattribution cosine similarityHurt
replace GraphSVX with TN-SHAP-G for first-order Shapley on small graphsBenzene · BenzeneO1 cosine similarity: 0.30 -> 0.9979Improved
replace GraphSVX with TN-SHAP-G for first-order Shapley on small graphsMutagenicity · MutagenicityO1 cosine similarity: 0.45 -> 0.9933Improved
use TN-SHAP-G instead of permutation sampling and SHAP-IQ at matched budgetsmolecular benchmarks · Mutagenicity, Benzene, PROTEINSqueries to reach 0.95 cosine: 10-100x fewer queriesImproved
amortize one trained surrogate over repeated explanation queries vs GraphSHAP-IQTN-SHAP-G · Mutagenicity (20 graphs, n<=20)total time: K=1: 197.8 s vs 11.6 s; K=100: 406.7 s vs 1158.1 sMixed
increase surrogate test R2 (more epochs/capacity)TN-SHAP-G · molecular graphsShapley cosine similarityImproved
use coarsening architecture to scale to larger graphsTN-SHAP-G · graphs of 100 to 600 nodestraining time scaling: T ~ n^2.30 training; T_Shapley ~ n^0.57Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy