AI Paper Atlas

Research map / Probabilistic and scientific modeling / Treatment effects

IGC-Net for conditional average potential outcome estimation over time

Konstantin Hess, Dennis Frauen, Valentyn Melnychuk, Stefan Feuerriegel

ICLR 2026 · Probabilistic and scientific modeling › Treatment effects · approach: Benchmarking treatment effect models · 9 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 IGC-Net instead of best baseline under high confoundingIGC-Net · Tumor growth synthetic (gamma=20)RMSE: 4.24 -> 3.71Improved
use IGC-Net on high-dimensional MIMIC-III semi-synthetic dataIGC-Net · MIMIC-III semi-synthetic (N=3000, tau=6)RMSE: 0.60 -> 0.48Improved
iterative G-computation instead of inverse propensity weighting (RMSNs)IGC-Net · MIMIC-III semi-synthetic (N=1000, tau=6)RMSE: 1.02 -> 0.54Improved
regression-based G-computation instead of distribution estimation (G-Net)IGC-Net · MIMIC-III semi-synthetic (N=1000, tau=6)RMSE: 1.03 -> 0.54Improved
use IGC-Net on real-world factual predictionIGC-Net · MIMIC-III real-world factualRMSE: 9.14 vs CT 9.05 (N=3000, tau=2)No clear effect
add iterative generation and learning steps (vs biased transformer without adjustment) (ablation)IGC-Net · semi-syntheticRMSEImproved
replace multi-input transformer with LSTM backbone (ablation)IGC-Net · semi-syntheticRMSENo clear effect
IGC-Net under severe overlap violation (rho 0.6)IGC-Net · Tumor growth overlap sensitivityRMSE: 2.78 vs CT 2.72Mixed

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