AI Paper Atlas

Research map / Probabilistic and scientific modeling / Treatment effects

Adjusting Prediction Model Through Wasserstein Geodesic for Causal Inference

Yuguang Yan, Haolin Yang, Zecong Chen, Weilin Chen, Ruichu Cai, Zhifeng Hao

ICLR 2026 · Probabilistic and scientific modeling › Treatment effects · approach: Latent-variable causal inference

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
add optimal-transport geodesic intermediate groups (ablation)G-learner · Jobs (out-sample)R_Pol: 0.26 -> 0.17Improved
add optimal-transport geodesic intermediate groups (ablation)G-learner · Jobs (out-sample)ε_ATT: 0.09 -> 0.06Mixed
add optimal-transport geodesic intermediate groups (ablation)G-learner · News (out-sample)ε_ATE: 0.28 -> 0.25Improved
add optimal-transport geodesic intermediate groups (ablation)G-learner · News (out-sample)√ε_PEHE: 2.85 -> 2.87No clear effect
add data filtering (ablation)G-learner · Jobs (out-sample)R_Pol: 0.19 -> 0.17Improved
add data filtering (ablation)G-learner · Jobs (out-sample)ε_ATT: 0.19 -> 0.06Improved
add data filtering (ablation)G-learner · News (out-sample)ε_ATE: 0.29 -> 0.25Improved
add data filtering (ablation)G-learner · Twins (out-sample)√ε_PEHE: 0.32 -> 0.32No clear effect
use KDE-estimated marginals instead of uniform in OT (ablation)G-learner · Jobs (out-sample)ε_ATT: 0.06 -> 0.23Hurt
use KDE-estimated marginals instead of uniform in OT (ablation)G-learner · News (out-sample)√ε_PEHE: 2.87 -> 2.87No clear effect
use G-learner instead of DragonNetDragonNet · Simulated (out-sample, mc = 1.4)√ε_PEHE: 0.46 -> 0.36Improved
use G-learner instead of ESCFRESCFR · Simulated (out-sample, mc = 1.4)√ε_PEHE: 0.96 -> 0.36Improved

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