arXiv:2603. 19186v3 Announce Type: replace Abstract: Randomized controlled trials (RCTs) are the gold standard for estimating treatment effects, yet they are often underpowered for detecting effect heterogeneity.
The paper proposes supervised learning objectives for estimating causal effects of high‑dimensional treatments, such as text or continuous variables, by recasting causal inference as a learning problem under realizability and no unobserved confounding. It shows that causal error can be decomposed into moment‑balancing errors of increasing order and designs objectives that directly improve causal estimation by exploiting the structure of conditional outcomes. The authors also demonstrate how to project high‑dimensional treatment effects onto coarser attributes, enabling a single model to answer multiple causal questions without extra attribute‑specific training, and validate their approach empirically across continuous, discrete, and text treatments.
arXiv:2507. 20993v4 Announce Type: replace-cross Abstract: We study how to learn treatment policies from multimodal electronic health records (EHRs) that consist of tabular data and clinical text.
The paper introduces new one‑step and targeted minimum loss‑based estimators for the average treatment effect and the average treatment effect on the treated under the front‑door criterion. These estimators use multiple parameterizations of the observed data distribution, can avoid modeling the mediator density, and are compatible with flexible machine‑learning nuisance estimation. The authors derive conditions for root‑n consistency, asymptotic linearity, and develop doubly robust tests for identification assumptions, demonstrating improved efficiency and favorable performance in simulations and real‑world applications in education and emergency medicine.
arXiv:2604. 23904v3 Announce Type: replace-cross Abstract: Synthetic tabular data are often evaluated by distributional similarity, privacy distance, or train-on-synthetic-test-on-real predictive performance, but these criteria do not ensure validity for causal inference.
ACTIVA is a transformer‑based conditional variational autoencoder designed to estimate full interventional distributions from observational data without requiring interventional data, a causal graph, or per‑task fitting. It learns a conditional latent prior that amortizes causal knowledge across diverse training tasks, enabling a single forward pass to handle new causal models drawn from the training distribution. Empirical results on synthetic and biologically realistic gene‑expression simulations show that ACTIVA outperforms correlational baselines, reduces spurious non‑descendant effects, and matches strong amortized baselines, while also validating its theoretical mixture‑of‑distributions property on observationally equivalent models.
arXiv:2609.17238v1 Announce Type: cross Abstract: High-dimensional data create challenges for causal effect estimation because identifying the covariates needed for correct model specification become...
The paper introduces a model‑agnostic inference framework for partially identified causal effects that leverages covariate information without requiring discrete covariates or accurate conditional distribution estimates. Using duality theory for optimal transport, the method delivers uniformly valid inference in randomized experiments, is doubly robust in observational settings, achieves asymptotic unbiasedness when nuisance parameters converge semiparametrically, and allows multiplier‑bootstrap selection of covariates and models while remaining computationally efficient. Empirical applications show the approach consistently narrows identified sets and confidence intervals without imposing extra structural assumptions.
arXiv:2604. 16763v3 Announce Type: replace Abstract: Causal inference from electronic health records (EHR) is fundamentally limited by unmeasured confounding: critical clinical states such as frailty, goals of care, and mental status are documented in free-text notes but absent from structured data.
Estimating heterogeneous treatment effects is central to targeted interventions, such as personalized promotions and precision medicine. We focus on the conditional average treatment effect (CATE), a standard estimand for characterizing such heterogeneity.
arXiv:2607. 26599v1 Announce Type: new Abstract: Estimating heterogeneous treatment effects is central to targeted interventions, such as personalized promotions and precision medicine.
The paper investigates how learned multimodal representations of high‑dimensional health data can fail to preserve causal information, even when trained to reconstruct the data. It establishes conditions under which a representation is sufficient for causal identification, shows that interventional mediation requires stricter preservation than counterfactual comparison, and proves that reconstruction‑based objectives cannot certify sufficiency. The authors decompose the error of cross‑fitted doubly robust estimators into a first‑order representation bias and a second‑order nuisance‑estimation remainder, demonstrating that coverage can deteriorate with larger samples if the representation is insufficient. They propose a practical workflow that combines modality‑specific variational autoencoders, cross‑fitted doubly robust estimation, locked phenotypes, selective pathway analysis, portable representations, and diagnostics for representation insufficiency, exposure leakage, and sensitivity, and validate it in simulations where a rare confounder’s omission leads to severely under‑covered indirect‑effect intervals.