Research map / Probabilistic and scientific modeling / Causal discovery
DECAF: Generating Fair Synthetic Data Using Causally-Aware Generative Networks
Boris van Breugel, Trent Kyono, Jeroen Berrevoets, Mihaela van der Schaar
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 |
|---|---|---|---|
| Remove FTU-violating edges at inference (ablation) | DECAF · Adult | FTU: 0.152±0.013 -> 0.004±0.004 | Improved |
| Remove DP-violating edges at inference (ablation) | DECAF · Adult | DP: 0.198±0.013 -> 0.001±0.001 | Improved |
| Apply DP edge removal rather than no debiasing (ablation) | DECAF · Adult | AUROC: 0.781±0.007 -> 0.672±0.014 | Hurt |
| Remove FTU-violating edges at inference (ablation) | DECAF · Credit Approval | AUROC: 0.713±0.045 -> 0.877±0.009 | Improved |
| Remove DP-violating edges at inference (ablation) | DECAF · Credit Approval | DP: 0.511±0.130 -> 0.011±0.029 | Improved |
| Use a fixed advantaged-group surrogate instead of sampling the parent marginal (ablation) | DECAF-FTU · Credit Approval | precision | Improved |
| Use a fixed advantaged-group surrogate instead of sampling the parent marginal (ablation) | DECAF-FTU · Credit Approval | FTU | No clear effect |
| Remove the protected attribute before synthesis (ablation) | DECAF · Adult | AUROC: 0.781±0.007 -> 0.769±0.020 | Hurt |
| Remove the protected attribute before synthesis (ablation) | DECAF · Adult | FTU: 0.152±0.013 -> 0.0±0.0 | Improved |
Closest papers by meaning
- AIM-Fair: Advancing Algorithmic Fairness via Selectively Fine-Tuning Biased Models with Contextual Synthetic Data (CVPR 2025 · similarity 0.89)
- FairPFN: A Tabular Foundation Model for Causal Fairness (ICML 2025 · similarity 0.86)
- CuTS: Customizable Tabular Synthetic Data Generation (ICML 2024 · similarity 0.86)
- FairWire: Fair Graph Generation (NeurIPS 2024 · similarity 0.86)
- Long-Term Fair Decision Making through Deep Generative Models (AAAI 2024 · similarity 0.85)
- Causal Bias Detection in Generative Artifical Intelligence (NeurIPS 2026 · similarity 0.85)
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.
