Local privacy mechanisms preserve rate-double-robustness, enabling unbiased and semiparametrically efficient inference on target parameters indexed linearly by infinite-dimensional and nonlinearly by low-dimensional components from noisy private data.
hub Mixed citations
Double/debiased machine learning for treatment and structural parameters
Mixed citation behavior. Most common role is method (60%).
42
Pith papers citing it
2,136
external citations · Crossref
Method
60% of classified citations
hub tools
citation-role summary
method 3
background 2
citation-polarity summary
representative citing papers
Sample-split random multipliers on held-out Fréchet losses yield asymptotically valid, nondegenerate tests of global/partial significance and global Fréchet specification for metric-space responses.
Develops stochastic policies and single-basis-function modification for causal inference on functional treatments, proves asymptotic normality and rate double robustness, and applies to NHANES physical activity and mortality data.
Introduces Trajectory Proper Score (TPS) as a strictly proper family of trajectory-level scoring rules that elicits the complete prefix-conditioned success probability process.
Introduces graph-to-image prediction of per-node dynamic stability landscapes in oscillator networks from topology, releases two 10k-graph datasets, and shows GNN-CNN models achieve good accuracy with cross-size generalization.
Algometrics proves that deployment risk cannot be identified from passive historical data alone, that model rankings can invert under crowding, and that randomized actions can identify short-horizon linear feedback.
DR-ME is the first semiparametrically efficient finite-location kernel test for interpretable distributional treatment effects, using orthogonal doubly robust features derived from observational data.
A methodological framework for separable effects analysis that distinguishes four-arm and two-arm designs, with EIF-based estimation and falsification tests.
PUICL is a transformer pretrained on synthetic PU data from structural causal models that solves positive-unlabeled classification via in-context learning without gradient updates or fitting.
AOI approximately inverts the likelihood mapping from fixed effects to outcomes to produce an estimator whose bias vanishes exponentially in T with double robustness.
Derives the efficient influence function and doubly robust estimators for the local average treatment effect on the treated in instrumented DiD designs with staggered exposure and covariates.
A GEE-based stacked M-estimation framework merges propensity score and marginal structural models to directly compute the large-sample variance of the IPTW estimator from pilot data for prospective sample size planning, with bootstrap stabilization.
Bitcoin fees are estimated via a Vickrey-Clarke-Groves model of mempool blockspace where congestion drives expected confirmation delay and the marginal value of priority is directly priced into chosen fees.
A unique observed-data functional for direct effects that equals the ATE if the focal variable is pre-exposure and an interventional direct effect if post-exposure, under a no-additive-interaction condition.
Staggered Medicaid doula mandates have no average LBW effect but a marginal ~0.5pp reduction for Black mothers in early-adopting states, with a strong first-stage workforce expansion.
A doubly cross-fit DR sieve learner identifies and estimates within-stratum heterogeneous treatment effects under principal ignorability and odds-ratio sensitivity, with oracle rates and uniform bands.
Heat-kernel smoothing over weighted points on a compact manifold yields a scale-dependent geometric effective sample size that discounts nearby and duplicate particles.
Non-closing Lie brackets of intervention-response fields can serve as a high-recall screen for causal edges under latent confounding, but do not by themselves identify general DAGs.
A new estimand and semiparametric inference framework measures treatment-effect heterogeneity across baseline-predicted post-treatment response profiles, avoiding latent principal strata.
Formalizes support-exclusion in constrained pricing and introduces target-aware controller with certified bands and regret-information accounting that identifies when polynomial target mass succeeds but 1/t branches fail without extra movement.
OPAL learns optimal smooth labeling policies from ML uncertainty scores to enable low-variance prediction-assisted inference with finite-sample coverage guarantees.
CausalGuard aggregates LLM-proposed and data-pruned DAGs to weight doubly robust pseudo-outcomes and applies conformal calibration to deliver finite-sample marginal coverage for conditional average treatment effects under graph uncertainty.
Double/debiased ML framework for average derivative effects in panel data with continuous treatments, two-way fixed effects, and endogeneity.
Develops Grenander-type and debiased machine learning estimators for the sublevel-set probability curve of the CATE function, shown to be non-pathwise differentiable, along with its piecewise linear approximation.
citing papers explorer
-
Private Rate-Double-Robust Inference
Local privacy mechanisms preserve rate-double-robustness, enabling unbiased and semiparametrically efficient inference on target parameters indexed linearly by infinite-dimensional and nonlinearly by low-dimensional components from noisy private data.
-
MATCH: Multiplier-Assisted Tests for Conditional Hypotheses in Non-Euclidean Data
Sample-split random multipliers on held-out Fréchet losses yield asymptotically valid, nondegenerate tests of global/partial significance and global Fréchet specification for metric-space responses.
-
Causal Inference for Functional Treatments with Stochastic Policies
Develops stochastic policies and single-basis-function modification for causal inference on functional treatments, proves asymptotic normality and rate double robustness, and applies to NHANES physical activity and mortality data.
-
Proper Scoring Rules for Agentic Uncertainty Quantification
Introduces Trajectory Proper Score (TPS) as a strictly proper family of trajectory-level scoring rules that elicits the complete prefix-conditioned success probability process.
-
Learning Dynamic Stability Landscapes in Synchronization Networks
Introduces graph-to-image prediction of per-node dynamic stability landscapes in oscillator networks from topology, releases two 10k-graph datasets, and shows GNN-CNN models achieve good accuracy with cross-size generalization.
-
Algometrics: Forecasting Under Algorithmic Feedback
Algometrics proves that deployment risk cannot be identified from passive historical data alone, that model rankings can invert under crowding, and that randomized actions can identify short-horizon linear feedback.
-
Semiparametric Efficient Test for Interpretable Distributional Treatment Effects
DR-ME is the first semiparametrically efficient finite-location kernel test for interpretable distributional treatment effects, using orthogonal doubly robust features derived from observational data.
-
Separable Effects in Four-Arm and Two-Arm Designs
A methodological framework for separable effects analysis that distinguishes four-arm and two-arm designs, with EIF-based estimation and falsification tests.
-
In-Context Positive-Unlabeled Learning
PUICL is a transformer pretrained on synthetic PU data from structural causal models that solves positive-unlabeled classification via in-context learning without gradient updates or fitting.
-
Approximate Operator Inversion for Average Effects in Nonlinear Panel Models
AOI approximately inverts the likelihood mapping from fixed effects to outcomes to produce an estimator whose bias vanishes exponentially in T with double robustness.
-
Doubly Robust Instrumented Difference-in-Differences
Derives the efficient influence function and doubly robust estimators for the local average treatment effect on the treated in instrumented DiD designs with staggered exposure and covariates.
-
Estimator-Aligned Prospective Sample Size Determination for Designs Using Inverse Probability of Treatment Weighting
A GEE-based stacked M-estimation framework merges propensity score and marginal structural models to directly compute the large-sample variance of the IPTW estimator from pilot data for prospective sample size planning, with bootstrap stabilization.
-
A Model and Estimation of the Bitcoin Transaction Fee
Bitcoin fees are estimated via a Vickrey-Clarke-Groves model of mempool blockspace where congestion drives expected confirmation delay and the marginal value of priority is directly priced into chosen fees.
-
Model-Robust Direct Effect Under Confounder-Mediator Ambiguity
A unique observed-data functional for direct effects that equals the ATE if the focal variable is pre-exposure and an interventional direct effect if post-exposure, under a no-additive-interaction condition.
-
Helping Hands, Healthier Infants: The Effect of Medicaid Doula Coverage Mandates on Birth Outcomes
Staggered Medicaid doula mandates have no average LBW effect but a marginal ~0.5pp reduction for Black mothers in early-adopting states, with a strong first-stage workforce expansion.
-
Doubly cross-fit debiased machine learning of heterogeneous treatment effects under principal stratification
A doubly cross-fit DR sieve learner identifies and estimates within-stratum heterogeneous treatment effects under principal ignorability and odds-ratio sensitivity, with oracle rates and uniform bands.
-
Heat-Kernel Entropy Profiles and Geometric Effective Sample Size for Weighted Measures on Manifolds
Heat-kernel smoothing over weighted points on a compact manifold yields a scale-dependent geometric effective sample size that discounts nearby and duplicate particles.
-
Latent Confounded Causal Discovery via Lie Bracket Geometry
Non-closing Lie brackets of intervention-response fields can serve as a high-recall screen for causal edges under latent confounding, but do not by themselves identify general DAGs.
-
Empirical stratification for predictive treatment effect heterogeneity with post-treatment variables
A new estimand and semiparametric inference framework measures treatment-effect heterogeneity across baseline-predicted post-treatment response profiles, avoiding latent principal strata.
-
Resource-Constrained Adaptive Inference for Sequential Pricing
Formalizes support-exclusion in constrained pricing and introduces target-aware controller with certified bands and regret-information accounting that identifies when polynomial target mass succeeds but 1/t branches fail without extra movement.
-
Optimized Labeling Resource Allocation for Prediction-Assisted Inference via OPAL
OPAL learns optimal smooth labeling policies from ML uncertainty scores to enable low-variance prediction-assisted inference with finite-sample coverage guarantees.
-
CausalGuard: Conformal Inference under Graph Uncertainty
CausalGuard aggregates LLM-proposed and data-pruned DAGs to weight doubly robust pseudo-outcomes and applies conformal calibration to deliver finite-sample marginal coverage for conditional average treatment effects under graph uncertainty.
-
Double/Debiased Machine Learning for Continuous Treatment Effects in Panel Data with Endogeneity
Double/debiased ML framework for average derivative effects in panel data with continuous treatments, two-way fixed effects, and endogeneity.
-
Nonparametric inference for sublevel-set probabilities of conditional average treatment effect functions
Develops Grenander-type and debiased machine learning estimators for the sublevel-set probability curve of the CATE function, shown to be non-pathwise differentiable, along with its piecewise linear approximation.
-
Causal Foundation Models with Continuous Treatments
A transformer meta-trained on a novel continuous-treatment data-generating prior reconstructs individual treatment-response curves from observational data via in-context learning and reports SOTA versus task-specific causal models.
-
Doubly Robust Proxy Causal Learning with Neural Mean Embeddings
A neural doubly robust proxy causal learning framework using mean embeddings for treatment bridges provides consistent estimators for causal dose-response functions under unobserved confounding for continuous and structured treatments.
-
CIVeX: Causal Intervention Verification for Language Agents
CIVeX maps agent tool calls to structural causal queries, checks identifiability, and issues auditable verdicts to prevent false executions while preserving utility on confounded benchmarks.
-
Improving Variance Estimation for Covariate Adjustment with Binary Outcomes
The IF-LOO variance estimator for covariate-adjusted treatment effects with binary outcomes provides appropriate type I error control in simulations, especially for rare events or small samples, with a closed-form implementation.
-
Inference on Linear Regressions with Two-Way Unobserved Heterogeneity
A Neyman-orthogonal moment estimator with adjusted nonparametric fixed effects achieves root-NT asymptotic normality for common parameters in two-way heterogeneous panel models.
-
Estimating heterogeneous treatment effects with survival outcomes via a deep survival learner
DSL uses doubly robust pseudo-outcomes and a multi-output neural network to jointly estimate time-varying conditional average treatment effects for right-censored survival data.
-
Targeted Regularization for Causal Effect Estimation with Exponential Dispersion Family Outcomes
Unified targeted regularization framework for causal effect estimation with EDF outcomes using neural networks that jointly estimate outcome model, propensity scores, and fluctuation parameter.
-
A Conformal Selection Framework for Individual Treatment Beneficiaries with Auxiliary External Data
A model-agnostic conformal selection method reformulates CATE-based beneficiary identification as multiple testing with RCT-calibrated p-values and FDR control, allowing external data for model training.
-
How to deal with machine learning bias in economic history
The paper guides ML use in economic history, identifies systematic prediction bias that distorts coefficients, and shows debiasing via small expert-labeled samples can correct it while preserving scale.
-
AutoResearchClaw: Self-Reinforcing Autonomous Research with Human-AI Collaboration
AutoResearchClaw introduces a multi-agent research pipeline with debate, self-healing, verifiable outputs, human collaboration modes, and cross-run evolution that outperforms AI Scientist v2 by 54.7% on ARC-Bench.
-
Estimating treatment duration effects via clone-censor-weight: a breast cancer case study
The clone-censor-weight approach is formalized and tested via simulations before application to a breast cancer cohort comparing 2 versus 5 years of adjuvant tamoxifen, yielding estimates with substantial uncertainty.
-
crossfit: A Graph-Based Cross-Fitting Engine in R
crossfit is an R package that supplies a general-purpose cross-fitting engine driven by user-specified DAGs of nuisance models with configurable fold allocations and reproducibility features.
-
BGM-IV: an AI-powered Bayesian generative modeling approach for instrumental variable analysis
BGM-IV performs nonlinear IV regression by inferring causally structured latent components and replacing the outcome likelihood with an instrument-averaged pseudo-likelihood, showing strongest results in high-dimensional covariate regimes.
-
TabCF: Distributional Control Function Estimation with Tabular Foundation Models
TabCF is a tuning-light method using tabular foundation models for control function regression to estimate distributional causal effects such as interventional means and quantiles.
-
RepFlow: Representation Enhanced Flow Matching for Causal Effect Estimation
RepFlow combines representation learning and conditional flow matching to estimate both point and distributional causal effects while mitigating selection bias via entropically regularized Wasserstein distance on normalized latent representations.
-
Identification of Latent Group Effects under Conditional Calibration
Under conditional calibration, a latent group effect is identified as a closed-form ratio of the residual covariance of the signed score with the outcome to twice the residual score variance.
-
Additive Control Variates Dominate Self-Normalisation in Off-Policy Evaluation
β*-IPS, the optimal additive control-variate estimator, asymptotically dominates SNIPS in MSE; the exact variance gap is (V(π)σ²_w − σ_{w,wr})²/(nσ²_w).
-
A Guide to Estimating Conditional Average Treatment Effects in Competing Risks Settings
Compares six meta-learners (Cox/RSF risk models paired with elastic net/RF CATE models) via simulations differing in hazard complexity and censoring, and releases the R package crsurvlearners.