Research map / Probabilistic and scientific modeling / Treatment effects
B-Learner: Quasi-Oracle Bounds on Heterogeneous Causal Effects Under Hidden Confounding
Miruna Oprescu, Jacob Dorn, Marah Ghoummaid, Andrew Jesson, Nathan Kallus, Uri Shalit
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 oracle versus estimated first-stage nuisances in the B-Learner pseudo-outcome (ablation) | B-Learner τ+(RF, RF) on simulated data · Simulated data (log Λ=1) | MSE vs true bound | No clear effect |
| Replace estimated-nuisance pseudo-outcome with plug-in estimator (no Neyman-orthogonal correction) (ablation) | B-Learner first stage on simulated data · Simulated data (log Λ=1) | MSE vs true bound | Hurt |
| Use NN or GK first stage with RF second stage instead of Sensitivity Kernel / Quince baselines | B-Learner · Semi-synthetic benchmark (simulated) | MSE | Improved |
| Use same nuisances as baselines (GK or NN) in B-Learner | B-Learner vs Sensitivity Kernel / Quince · Semi-synthetic benchmark (simulated) | MSE | No clear effect |
| RF B-Learner instead of GK-based Sensitivity Kernel | IHDP Hidden Confounding deferral · IHDP hidden confounding | recommendation error rate | Improved |
| NN B-Learner versus Quince | IHDP hidden confounding deferral · IHDP hidden confounding | recommendation error rate | No clear effect |
| Increase sensitivity parameter log Λ from 0 to 0.6 | 401(k) eligibility CATE lower bounds · 401(k) eligibility wealth data | fraction of negative lower bounds: about half negative at log Λ=0.6 | Hurt |
Closest papers by meaning
- A Meta-learner for Heterogeneous Effects in Difference-in-Differences (ICML 2025 · similarity 0.92)
- Do Contemporary Causal Inference Models Capture Real-World Heterogeneity? Findings from a Large-Scale Benchmark (ICLR 2025 · similarity 0.90)
- A Non-parametric Direct Learning Approach to Heterogeneous Treatment Effect Estimation under Unmeasured Confounding (NeurIPS 2024 · similarity 0.90)
- Quantifying Aleatoric Uncertainty of the Treatment Effect: A Novel Orthogonal Learner (NeurIPS 2024 · similarity 0.90)
- Estimating Heterogeneous Treatment Effects by Combining Weak Instruments and Observational Data (NeurIPS 2024 · similarity 0.89)
- Learning Representations of Instruments for Partial Identification of Treatment Effects (ICML 2025 · similarity 0.89)
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.
