Research map / Probabilistic and scientific modeling / Treatment effects
Observational-Interventional Priors for Dose-Response Learning
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| fix a=1, b=0 (drop the affine correction of f_obs) (ablation) | Proposed prior · Synthetic, adversarial confounding | mean error / log-likelihood | Hurt |
| fix a and b (Competitor I) uncertainty (ablation) | Proposed prior · Synthetic | log-likelihood | Hurt |
| replace affine transformation with generic SE-GP residual (Competitor II) (ablation) | Proposed prior · Synthetic, small M | mean error: 0.05 -> 0.02 | Hurt |
| use GP on interventional data only (Competitor III) (ablation) | Proposed prior · Synthetic, M=40 vs 200 | mean error: 0.11 -> 0.03 | Hurt |
| increase interventional sample size M from 40 to 200 | Proposed advantage over Competitor III · Synthetic Q50% Random | error gap: 0.11 -> 0.03 | Mixed |
| low signal-to-noise problems | Competitor III · Synthetic | log-likelihood | Hurt |
| add observational prior to GP on interventional data | GP regression · IHDP case study, high school stratum | normalized absolute error: 0.06 | Helped |
| add observational prior to GP on interventional data | GP regression · IHDP case study, college stratum | normalized absolute error: 0.07 | Helped |
| add observational prior to GP on interventional data | GP regression · IHDP case study, combined | normalized absolute error: 0.08 | Helped |
| larger signal range 80% vs 50% with cubic truth (ablation) | Competitor I gap · Synthetic cubic Random | mean error difference: 0.01 -> 0.03 | Mixed |
Closest papers by meaning
- Continuous Treatment Effect Estimation Using Gradient Interpolation and Kernel Smoothing (AAAI 2024 · similarity 0.85)
- Causal Identification for Complex Functional Longitudinal Studies (ICLR 2025 · similarity 0.85)
- Learning Representations of Instruments for Partial Identification of Treatment Effects (ICML 2025 · similarity 0.85)
- Causal Attribution Analysis for Continuous Outcomes (ICML 2025 · similarity 0.84)
- B-CALM: Bias-Limited Bayesian Borrowing for RCT-Anchored Treatment Effects under Covariate Mismatch (NeurIPS 2026 · similarity 0.84)
- Quantifying Aleatoric Uncertainty of the Treatment Effect: A Novel Orthogonal Learner (NeurIPS 2024 · 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.