Research map / Probabilistic and scientific modeling / Molecules and proteins
Deep Nonparametric Conditional Independence Tests for Images
Marco Simnacher, Xiangnan Xu, Hani Park, Christoph Lippert, Sonja Greven
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 |
|---|---|---|---|
| use RCoT as the nonparametric CIT instead of Wald test on embeddings | DNCIT · Simulation, nonlinear confounding, n over 1475 | type I error | Improved |
| use RCoT instead of CMIknn at large sample sizes | DNCIT · Simulation, n above 1100 | power and runtime | Improved |
| use CMIknn instead of RCoT at small sample sizes | DNCIT · Simulation, n up to 1100 | T1E and power | Improved |
| increase confounder dimension from 1 to 10 (ablation) | Deep-CPT-KPC · Simulation, Freesurfer embedding | power | Hurt |
| swap embedding map (Fastsurfer, noisy Fastsurfer, Freesurfer, cVAE) (ablation) | DNCIT · Simulation, T1E | type I error | No clear effect |
| swap embedding map (ablation) | DNCIT · Simulation, power | power | Mixed |
| vary confounder relationship from linear to squared to complex (ablation) | DNCIT · Simulation | power | No clear effect |
| use RCoT on whole brain MRI features instead of Wald test | brain structure vs personality traits (UKB, n=8634) · UK Biobank BFBT association | p-values | Mixed |
Closest papers by meaning
- Toward Scalable and Valid Conditional Independence Testing with Spectral Representations (ICML 2026 · similarity 0.87)
- Conditional Diffusion Models Based Conditional Independence Testing (AAAI 2025 · similarity 0.86)
- On the Hardness of Conditional Independence Testing In Practice (NeurIPS 2025 · similarity 0.85)
- A Conditional Independence Test in the Presence of Discretization (ICLR 2025 · similarity 0.84)
- PAIR-CI: Calibrated Conditional Independence Testing for Causal Discovery with Incomplete Data (NeurIPS 2026 · similarity 0.84)
- Testing Conditional Mean Independence Using Generative Neural Networks (ICML 2025 · similarity 0.84)
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.
