Research map / Probabilistic and scientific modeling / Treatment effects
Counterfactuals for the Future
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 |
|---|---|---|---|
| Choose unit-specific treatments using observed outcomes rather than a covariate-based EWM rule | Four-unit SCM example · Four-unit SCM example | Gini social welfare: \frac{26}{16} -> \frac{32}{16} | Helped |
| Use a forward-looking counterfactual estimate that reuses unit-specific exogenous noise | Estimating post-intervention outcomes when σ_U=5 and σ_μ=5 · σ_U=5, σ_μ=5, δ=1 | variance error relative to the true distribution: 9.61 versus 9.36; interventional estimate: 51.1 | Helped |
| Use an interventional estimate rather than a forward-looking counterfactual estimate when exogenous noise varies over time and unit structure is sufficiently small | Estimating the future outcome distribution · Some settings with σ_U>0 and sufficiently small σ_μ | KL divergence to the true outcome distribution | Mixed |
| Use forward-looking counterfactuals when unit-specific exogenous factors are sufficiently dissimilar | Estimating the future outcome distribution · Increasing σ_μ | KL divergence to the true outcome distribution | Helped |
| Use forward-looking counterfactuals when exogenous factors remain constant within each unit over time | Estimating the future outcome distribution · σ_U=0 and σ_μ>0 | distributional match to the true outcome | Helped |
Closest papers by meaning
- Learning Counterfactual Outcomes Under Rank Preservation (NeurIPS 2025 · similarity 0.86)
- An Empirical Examination of Balancing Strategy for Counterfactual Estimation on Time Series (ICML 2024 · similarity 0.85)
- Counterfactual Realizability (ICLR 2025 · similarity 0.85)
- Towards Characterizing Domain Counterfactuals for Invertible Latent Causal Models (ICLR 2024 · similarity 0.84)
- Counterfactual Contrastive Learning with Normalizing Flows for Robust Treatment Effect Estimation (ICML 2025 · similarity 0.84)
- Cyclic Counterfactuals under Shift–Scale Interventions (NeurIPS 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.