Research map / Learning theory and algorithms / Optimal transport
Optimal Transport Group Counterfactual Explanations
Enrique Valero-Leal, Bernd Bischl, Pedro Larrañaga, Concha Bielza, Giuseppe Casalicchio
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 changed | Where | Measured | Effect |
|---|---|---|---|
| replace pointwise group counterfactuals with an explicit affine/Gaussian OT map under strict bi-Lipschitz (K=1.01) | OT group counterfactuals vs Group w/ bi-Lipschitz · 15 tabular datasets, K=1.01 | convergence: baseline fails to converge in 50% of experiments | Improved |
| use OT maps instead of baseline across K values | OT group counterfactuals · 15 tabular datasets, all K | convergence: baseline fails in 35% | Improved |
| enforce bi-Lipschitz via convex SDP/QCQP maps instead of unconstrained Wachter | PSD affine and others · W2 transport cost | W2: 30% higher W2 than unconstrained Wachter | Hurt |
| use larger K (looser bi-Lipschitz) (ablation) | baseline vs OT maps · W2 gap | W2 | Mixed |
| use dense maps (PSD affine, Gaussian) instead of sparse maps (ablation) | OT maps · W2 transport cost | W2: 90% within factor 1.7 vs 80% within 2.3 | Improved |
| use sparse maps for efficiency (W2 x log params) (ablation) | OT maps · efficiency | efficiency: baseline factor 2.2 vs ours 1.1 to 1.9 | Improved |
| use sparse maps in multiobjective NSGA-II optimization (ablation) | OT maps · hypervolume | hypervolume: diag affine ~1.01, scaled Gaussian ~1.4 ratio | Improved |
| use scaled Gaussian transform (ablation) | Gaussian scaled · counterfactual validity | validity: validity near 1 for all K | Improved |
| add the Lipschitz upper bound constraint (ablation) | Group counterfactual · W2 and distortion | upper bound | No clear effect |
| use OT maps instead of baseline for runtime | OT group counterfactuals · runtime | runtime: baseline ~1250x (Lipschitz) and ~60000x (bi-Lipschitz) slower than fastest proposal | Improved |
Closest papers by meaning
- Counterfactual Maps: What They Are and How to Find Them (NeurIPS 2026 · similarity 0.88)
- Counterfactual Metarules for Local and Global Recourse (ICML 2024 · similarity 0.85)
- Collective Counterfactual Explanations: Balancing Individual Goals and Collective Dynamics (NeurIPS 2025 · similarity 0.84)
- Model Reconstruction Using Counterfactual Explanations: A Perspective From Polytope Theory (NeurIPS 2024 · similarity 0.84)
- ExMap: Leveraging Explainability Heatmaps for Unsupervised Group Robustness to Spurious Correlations (CVPR 2024 · similarity 0.83)
- LeapFactual: Reliable Visual Counterfactual Explanation Using Conditional Flow Matching (NeurIPS 2025 · similarity 0.83)
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.
