Spatial Deconfounder: Interference-Aware Deconfounding for Spatial Causal Inference

Ai Kagawa; Ayush Khot; Maresa Schr\"oder; Miruna Oprescu; Xihaier Luo

REVIEW 3 major objections 4 minor 1 cited by

Spatial Deconfounder: Interference-Aware Deconfounding for Spatial Causal Inference

T0 review · 3 major / 4 minor · reviewed 2026-08-04 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read The paper claims that interference between neighboring sites exposes the latent spatial confounders that otherwise bias estimates, making direct and spillover effects identifiable from local treatment vectors alone.

desk verdict Genuinely interesting idea—interference as the multi-cause signal—but the identification proof doesn't hold together: the key cancellation requires ∇_a g = 0, which the implemented CVAE doesn't satisfy and piecewise constancy doesn't imply. read the letter →

arxiv 2510.08762 v2 pith:C7VZ4CJP submitted 2025-10-09 cs.LG stat.ML

Ayush Khot , Miruna Oprescu , Maresa Schröder , Ai Kagawa , Xihaier Luo This is my paper

classification cs.LGstat.ML

keywords spatialcausalinferenceinterferenceunobservedconfoundingdeconfoundingconditionalvariationalautoencoderdirecteffectsspilloveridentification

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper's central claim is that in spatial settings, unmeasured confounding and interference are not separate nuisances: interference itself creates a multi-cause structure that reveals the hidden confounder. The authors propose a two-stage procedure: a conditional variational autoencoder with a spatial smoothing prior reconstructs a substitute confounder from each site's own treatment plus its neighbors' treatments; then a flexible spatial outcome model estimates direct and spillover effects via plug-in contrasts. They prove a causal identifiability theorem showing these contrasts are valid without requiring a known parametric latent-field model or multiple treatment types, and they demonstrate on semi-synthetic environmental and social data that the method reduces bias relative to spatial autoregressive, spline, matching, and graph-convolution baselines. A sympathetic reader would care because it offers a route to valid effect estimates in observational spatial studies where both hidden drivers and spillovers are the norm, not the exception.

What carries the argument

The substitute confounder Z learned by a conditional variational autoencoder with a Gaussian Markov random field prior; the identifying contrasts are the plug-in differences E[Y | A_s=1, A_neighbors, X, Z] − E[Y | A_s=0, A_neighbors, X, Z] and the analogous spillover contrasts. Z summarizes the latent spatial field from the local multi-cause treatment vector, and the piecewise-constant assumption makes Z's outcome contribution cancel, converting unobservable confounding into an observable conditioning variable.

What would settle it

Run a simulation with a known smooth latent field U and an outcome whose true Z-dependence varies with treatment (i.e., where the encoder's Z is a smooth, non-piecewise-constant function of the treatment): train the proposed CVAE, compute the plug-in contrast, and compare it to the known true direct effect; if the bias persists with infinite data or the gradient of Z with respect to the treatment input is nonzero, the identifiability claim as stated fails.

Watch

Extended reading notes

Core claim

The central discovery is that the joint distribution of a site's treatment and its neighbors' treatments acts as a multi-cause proxy for the unobserved spatial field that confounds them, so a latent-variable model trained on those local treatment vectors can recover a substitute confounder Z. Once Z is included as a conditioning variable, the paper argues that ignorability holds for the joint exposure (own plus neighborhood treatment), and by Theorem 1 the average direct and spillover effects are identified as plug-in contrasts of the observed outcome regression E[Y | A_s, A_neighbors, X, Z]. Identification does not require multiple treatment types or a specified latent-field model; it requi

Load-bearing premise

The identification proof hinges on the learned substitute confounder being a deterministic, piecewise-constant function of the assigned treatments and covariates, so that changing treatment leaves its additive effect on the outcome unchanged; a smooth neural encoder will generally violate this condition, and without it the Z-dependent terms do not cancel.

Editorial extensions

If this is right

Direct and spillover effects can be estimated in observational spatial studies without a specified latent-field model, as long as local treatment vectors are observed.
The approach lets researchers combine deconfounding with flexible outcome models such as convolutional or graph-based architectures, extending nonparametric causal inference to gridded environmental and social data.
In contrast to many existing spatial baselines, the method also provides estimates of spillover effects, not just direct effects.
Because the substitute confounder is recovered from observable treatments and covariates, unmeasured spatial confounding need not be fatal if there is at least local interference.
The paper reports consistent bias reduction relative to spatial autoregressive, spline, matching, and graph-convolution baselines across several semi-synthetic datasets.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

The identification result depends on a cancellation that requires the substitute confounder to be a piecewise-constant function of the treatment; a smoothly trained neural encoder will generally not satisfy this, so it would be worth testing the method on data where the true latent field induces a smooth, nonzero-gradient dependence of Z on treatment — under such conditions, the same contrasts may
The assumption that the encoder recovers a bijective transformation of the true latent field is untestable from the observed data; practitioners may want to pair this method with sensitivity analyses for the strength of that assumption.
A testable extension follows: if the core claim is right, adding more distinct treatment types or richer neighborhood structures should improve confounder reconstruction, which could be verified by measuring how well the recovered Z predicts a held-out treatment variable.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

Short version: the paper's central idea is genuinely worth discussing, and the benchmark extension is useful, but Theorem 1's proof doesn't establish what it claims. The cancellation at Eqs. (42)–(43) requires ∇_a g = 0, and that condition is neither true for a neural CVAE encoder nor implied by the theorem's piecewise-constancy assumption. Even if you grant piecewise constancy, a function can jump between A_s=0 and A_s=1 while having zero derivative almost everywhere, so the finite-difference contrasts E[f2(g(1))]−E[f2(g(0))] don't vanish. The proof conflates an a.e. zero derivative with equality of endpoint values. That cancellation is the load-bearing step that removes the unobserved confounder, so the identification result as stated is unsupported.

What's good: framing interference as the multi-cause signal is a real reframing of the deconfounder idea for spatial grids. The two-stage CVAE-with-spatial-prior plus outcome model is sensible, and extending SpaCE to include interference is a useful service. The experiments cover several real-data environments with semi-synthetic outcomes, and the method often does reduce bias—though not always. In some appendix configurations classical baselines achieve lower direct-effect bias, and the spillover comparison is only against UNET because the other baselines don't produce spillover estimates. So the empirical claims in the abstract are a bit stronger than the tables show.

The other soft spot is Assumption 5. It basically asserts that the CVAE learns a bijective transformation of the true latent field. That's close to assuming the conclusion—if you already have the true confounder up to reparameterization and the outcome contribution is separable, identification is hardly surprising. The theorem's extra piecewise-constancy condition is doing the work, but it's not satisfied by the actual encoder and isn't motivated by the training objective.

My read: this paper deserves a serious referee, because the novelty is real and the empirical setup is useful, but the identification theorem as written should not be accepted. A corrected proof might use finite differences with an explicit condition like E[f2(g(1))]=E[f2(g(0))], or the authors could weaken the paper's claims and present the method as a practical heuristic. Either way, the current version needs substantial revision before the central claim is credible.

Referee Report

3 major / 4 minor

Summary. The paper proposes the Spatial Deconfounder, a two-stage method for estimating direct and spillover causal effects in spatial data under unobserved spatial confounding and interference. Stage 1 fits a conditional variational autoencoder (CVAE) with a spatial prior to the vector of own and neighborhood treatments, using the latent variable as a substitute confounder. Stage 2 fits a flexible outcome model conditional on the substitute confounder and estimates effects by plug-in contrasts. The central theoretical claim (Theorem 1) is nonparametric identification of direct and spillover effects under Assumptions 1–5. The paper also extends the SpaCE benchmark to include interference and reports empirical improvements over several spatial baselines. The main advertised contribution is turning interference from a nuisance into a multi-cause identifying signal for hidden spatial confounding.

Significance. If the identification result were sound, the paper would make a valuable conceptual contribution: it connects the multi-cause deconfounding idea of Wang and Blei (2019) to spatial interference, potentially enabling point identification in a setting where existing methods require strong parametric priors, IV exclusion restrictions, or full ignorability. The empirical study extends a public benchmark, and the authors provide replication code. These are genuine strengths. However, the central theorem's proof contains a load-bearing gap: the cancellation of the confounder-dependent terms relies on a derivative condition that is not satisfied by the implemented estimator and is insufficient for the binary-treatment contrasts actually used. Because the identification claim is the paper's core theoretical contribution, this issue is decisive for the current version.

major comments (3)

[Appendix B.2, Eqs. (42)–(43) and (49)–(50)] The proof's key cancellation step is invalid as written. The paper asserts that ∇_{a_s} E[f2(g)] = ∇_{a_s} E[f4(g)] = 0 because Z is piecewise constant, and then uses this to drop f2 and f4 from the final contrasts. But A_s is binary: the estimand is a finite difference g(1,a_N,x) − g(0,a_N,x), not a derivative. A piecewise-constant g can have ∇_{a_s} g = 0 almost everywhere while its endpoint values differ, so E[f4(g(1,a_N,X))] − E[f4(g(0,a_N,X))] need not vanish. The transition from Eqs. (48)–(50) (and analogously Eqs. (54)–(56)) assumes exactly this endpoint equality, which does not follow from a.e. zero derivative. Thus the confounder-dependent offset is not shown to cancel, and the identification conclusion of Theorem 1 is not established by the given proof.
[§5, Theorem 1 vs. §4, Eq. (6)] The theorem's hypotheses are not satisfied by the method being proposed. Theorem 1 requires Z to be a piecewise-constant function of the assigned causes and covariates, but the implemented encoder in Eq. (6) is a smooth neural network q_ϕ(Z_s | A_s, A_Ns, X_s, X_Ns) = N(µ_ϕ, diag σ_ϕ²), and the training objective (Eq. (8)) does not encourage piecewise constancy. Moreover, Assumption 5 asserts that the encoder converges to a degenerate deterministic function that is a bijective transformation of the true latent field U(s); this is a very strong, untestable condition, and the abstract's characterization of the assumptions as 'weak' is therefore overstated. Because the proof's cancellation relies on this piecewise-constant condition, the identification theorem does not cover the CVAE-based procedure actually used in the experiments.
[§5, Eqs. (27)–(30) and proof line (42)] The proof represents direct and spillover effects as path integrals over continuous variables ν ∈ R and κ ∈ R^{|S|−1} (Eqs. (27)–(30)). This is not justified for binary treatments: the contrast τ_dir is defined at A_s = 1 versus 0, not along a continuous path. Even if one introduced a continuous extension of f1 and f3, the equality of gradients in (44)–(45) is only an almost-everywhere statement, and the fundamental theorem of calculus cannot be applied unless the functions are absolutely continuous and the endpoints lie in the set where the derivative equality holds. The proof gives no such regularity argument. This is not merely a matter of notation; it is the mechanism by which the unobserved-confounder terms are removed, so the identifiability claim depends on it.

minor comments (4)

[§3, Assumption 3] Assumption 3 is stated as 'Localized interference' with no formal content; the actual localized-interference assumption appears in Eq. (2). Please either state the assumption explicitly or remove the numbered assumption.
[Algorithm 1, line 4 and Eq. (8)] The KL divergence is written as D_KL(q_ϕ ∥ p_ψ), but the prior is p_θ(Z) = N(0, τ^{-1}(L+εI)^{-1}). The subscript on the prior distribution should be θ, not ψ, or the notation should be clarified.
[Tables 1 and 2] Most baselines report 'N/A' for spillover effects, so the empirical comparison for the spillover estimand is effectively only between CVAE variants and UNET. This limits the strength of the claim that the method 'consistently improves effect estimation' for spillover effects. Please discuss this limitation explicitly.
[General notation] The paper switches between 'τ_ind' and 'τ_spill' for the spillover effect; please use one term consistently. Also, Eq. (10) is referenced from Algorithm 1 before it is defined; reorder or adjust the cross-reference.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the identification claim is conditional on explicit theoretical assumptions; the ∇g cancellation issue flagged by the reviewer is a proof-gap/correctness concern, not a circular reduction.

full rationale

The paper's derivation chain is conditional rather than circular. Theorem 1 states identifiability of direct and spillover effects under Assumptions 1–5 plus an explicit separability condition and a piecewise-constancy condition on the substitute confounder Z. These are stated assumptions, not fitted values, and the theorem does not define the causal estimand in terms of the fitted CVAE output. The CVAE in Stage 1 is trained on treatments and covariates only; outcomes enter only in Stage 2, so the reported effect estimates are not fitted to the target effects by construction. Lemma 1, which supplies ignorability given the substitute confounder, is imported from Wang and Blei (2019), an external published result rather than a self-citation by the present authors; no load-bearing self-citation chain appears. The skeptic's concern about Eqs. (42)–(50) is a substantive mathematical objection: the proof cancels f2/f4 terms by asserting ∇g=0 from piecewise constancy, but for binary treatments the relevant contrast is g(1)−g(0), which need not vanish, and the smooth CVAE encoder does not imply piecewise constancy. That is a proof gap / violation of the stated assumptions, not a circularity: the theorem does not assume the conclusion, and the identification conclusion is not equivalent to an input by construction. The paper's own Limitations section also concedes that the theoretical guarantees rely on idealized assumptions such as localized interference and smooth latent confounders. Therefore, under the circularity standards of this review, no circular step is established.

Assumptions & free parameters 3 free parameters · 7 assumptions · 1 invented entities

The ledger shows the theoretical claim depends on importing Wang-Blei's deconfounder identification (Prop 5) and on two paper-specific postulates: Assumption 5 (the learned Z is a bijective transformation of the true latent field) and the theorem's separability/piecewise-constant conditions. No free parameters enter the theorem, but the empirical headline depends on tuned choices d_Z, R, β, λ. The invented entity Z has no independent falsifiable handle.

free parameters (3)

CVAE latent dimension d_Z = 1-32, tuned per dataset
Selected by Ray Tune; causal estimates change materially with d_Z (Appendix Tables 5-6 report different biases for different d_Z for the same setting).
Deconfounder neighborhood radius R = 0, 1, or 2
Set by the analyst; the reported spillover bias changes from 0.31 ± 0.08 (R=1) to 1.47 ± 1.91 (R=2) in Table 1, showing the headline results depend on this choice.
CVAE KL weight β and spatial-prior penalties λ_t, λ_y = tuned (Table 4)
Regularization strengths in the CVAE and outcome module are tuned per dataset and shape the reconstructed substitute confounder.

assumptions (7)

domain assumption Assumption 1: Spatial consistency: observed outcome equals potential outcome under assigned own and neighbor treatments.
Standard consistency extended to localized interference; Section 5, Eq. (13).
domain assumption Assumption 2: Spatial positivity: 0 < P(a_s,a_Ns|X_s,X_Ns) < 1 and latent positivity given Z.
Standard overlap needed for plug-in contrasts; Section 5.
domain assumption Assumption 3: Localized interference: Y_s depends only on own and neighbor treatments.
Limits spillover radius; Section 3, Eq. (2).
domain assumption Assumption 4: Latent field sufficiency: no unobserved confounder acts on a single treatment only, and the conditional independence holds uniformly across sites.
Spatial analogue of single ignorability borrowed from Wang-Blei 2019; the formal statement in Section 5 is ambiguous (if it holds 'for some sites' then 'uniformly').
ad hoc to paper Assumption 5: The CVAE encoder converges to a degenerate deterministic function that is a bijective transformation of the true latent field U(s).
This is the key assumption that the learned Z actually captures the confounder; it is not derived from the CVAE objective and is not verified against outcomes.
ad hoc to paper Theorem 1: Outcome conditional expectations are additively separable (f1+f2, f3+f4) and Z is piecewise constant with vanishing treatment-gradient.
These conditions are introduced inside the theorem to make the proof's gradient cancellation work; they are not consequences of Assumptions 1-5 and are not satisfied by a smooth neural encoder.
standard math Wang and Blei (2019) Proposition 5: Under a factor-model representation and single ignorability, treatments are ignorable given the substitute confounder.
Imported theorem; the paper's Lemma 1 relies on it. Its validity is assumed without addressing the D'Amour critique.

invented entities (1)

Substitute confounder Z_s (CVAE latent variable)
purpose: Proxy for the unobserved spatial field U(s) to enable deconfounding
The only supporting evidence is a predictive check (Eqs. 11-12) showing the decoder reproduces treatment assignments; it does not verify that Z matches U or removes confounding. There is no external falsifiable handle.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Spatial Deconfounder: Interference-Aware Deconfounding for Spatial Causal Inference." pith.science (2026). https://pith.science/paper/C7VZ4CJP

@misc{pith2026251008762,
  author       = {Pith},
  title        = {Pith review of: Spatial Deconfounder: Interference-Aware Deconfounding for Spatial Causal Inference},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/C7VZ4CJP}},
  note         = {Machine review of arXiv:2510.08762}
}

read the original abstract

Causal inference in spatial domains faces two intertwined challenges: (1) unmeasured spatial factors, such as weather, air pollution, or mobility, that confound treatment and outcome, and (2) interference from nearby treatments that violate standard no-interference assumptions. While existing methods typically address one by assuming away the other, we show they are deeply connected: interference reveals structure in the latent confounder. Leveraging this insight, we propose the Spatial Deconfounder, a two-stage method that reconstructs a substitute confounder from local treatment vectors using a conditional variational autoencoder (C-VAE) with a spatial prior, then estimates causal effects with a flexible outcome model. We show that this enables nonparametric identification of direct and spillover effects under weak assumptions--without multiple treatment types or a known latent-field model. Empirically, we extend SpaCE, a benchmark suite for spatial confounding, to include treatment interference, and show that the Spatial Deconfounder consistently improves effect estimation across real-world environmental health and social science datasets. By turning local interference into a multi-cause proxy for latent spatial confounding, our framework advances robust causal inference for spatial data.

Figures

Figures reproduced from arXiv: 2510.08762 by the authors.

**Figure 1.** Schematic of spatial interference/confounding. Spatial data is represented in geographical cells indexed by site s with neighborhood Ns. The outcome at s (e.g., mortality rate) is affected by the treatments (e.g., air quality) and observed confounders (e.g., demographic informataion) at both s and Ns. However, unobserved latent factors(e.g., humidity) can confound the relationship, rendering causal effects unident… view at source ↗

**Figure 2.** Example spatial distribution of (normalized) confounder, treatment, and outcome in real [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗

**Figure 3.** Architecture of the spatial deconfounder & estimation framework. Stage 1 : The CVAE takes treatments and observed confounders as input to learn the latent substitute confounder. Stage 2 : We employ the reconstructed confounder together with the observed variables (now including the outcome) to train the potential outcome estimation module. 4 [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗

Discussion (0). Sign in to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

Spatiotemporal Proximal Causal Inference under Hidden Confounding and Interference
cs.LG 2026-08 conditional novelty 4.0 of 10

A spatiotemporal proximal causal inference framework identifies potential outcomes under hidden confounding and interference via a neighborhood-aware bridge function and a neural proxy learning pipeline.

Reference graph

Works this paper leans on

61 extracted references · 5 linked inside Pith · cited by 1 Pith paper

[1]

write newline

" write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 global.max substring 't := if while FUNCTION format.date year duplicate empty "emp...
[2]

S. Ali, O. Faruque, and J. Wang. Estimating direct and indirect causal effects of spatiotemporal interventions in presence of spatial interference. In Joint European Conference on Machine Learning and Knowledge Discovery in Databases, 2024

2024
[3]

Amini, M

H. Amini, M. Danesh-Yazdi, Q. Di, W. Requia, Y. Wei, Y. Abu-Awad, L. Shi, M. Franklin, C.-M. Kang, J. Wolfson, P. James, R. Habre, Q. Zhu, J. Apte, Z. Andersen, I. kloog, F. Dominici, P. Koutrakis, and J. Schwartz. Hyperlocal super-learned PM 2.5 components across the contiguous us, 2022

2022
[4]

J. D. Angrist, G. W. Imbens, and D. B. Rubin. Identification of causal effects using instrumental variables. Journal of the American Statistical Association, 91 0 (434): 0 444--455, 1996

1996
[5]

L. Anselin. Spatial Econometrics: Methods and Models, volume 4. Springer Science & Business Media, 1988

1988
[6]

P. M. Aronow and C. Samii. Estimating average causal effects under general interference, with application to a social network experiment. The Annals of Applied Statistics, 11 0 (4): 0 1912–--1947, 2017

1912
[7]

I. Bica, A. Alaa, and M. Van Der Schaar. Time series deconfounder: Estimating treatment effects over time in the presence of hidden confounders. In International conference on machine learning, 2020

2020
[8]

Behavioral risk factor surveillance system ( BRFSS ), 2010

Centers for Disease Control and Prevention . Behavioral risk factor surveillance system ( BRFSS ), 2010

2010

Show all 61 references

[9]

W. Chen, R. Cai, Z. Yang, J. Qiao, Y. Yan, Z. Li, and Z. Hao. Doubly robust causal effect estimation under networked interference via targeted learning. In Proceedings of the 41st International Conference on Machine Learning, 2024

2024
[10]

Cheng, R

L. Cheng, R. Guo, R. Moraffah, P. Sheth, K. S. Candan, and H. Liu. Evaluation Methods and Measures for Causal Learning Algorithms . IEEE Transactions on Artificial Intelligence, 3 0 (06): 0 924--943, 2022

2022
[11]

Curth, D

A. Curth, D. Svensson, J. Weatherall, and M. van der Schaar. Really doing great at estimating CATE ? a critical look at ML benchmarking practices in treatment effect estimation. In Advances Neural Information Processing Systems Datasets and Benchmarks Track (Round 2), 2021

2021
[12]

Deryugina, G

T. Deryugina, G. Heutel, N. H. Miller, D. Molitor, and J. Reif. The mortality and medical costs of air pollution: Evidence from changes in wind direction. American Economic Review, 109 0 (12): 0 4178--4219, 2019

2019
[13]

Q. Di, H. Amini, L. Shi, I. Kloog, R. Silvern, J. Kelly, M. B. Sabath, C. Choirat, P. Koutrakis, A. Lyapustin, Y. Wang, L. J. Mickley, and J. Schwartz. An ensemble-based model of pm2.5 concentration across the contiguous united states with high spatiotemporal resolution. Envir...

2019
[14]

J. Dorn, K. Guo, and N. Kallus. Doubly-valid/doubly-sharp sensitivity analysis for causal inference with unmeasured confounding. Journal of the American Statistical Association, 120 0 (549): 0 331--342, 2025

2025
[15]

Dosovitskiy, L

A. Dosovitskiy, L. Beyer, A. Kolesnikov, D. Weissenborn, X. Zhai, T. Unterthiner, M. Dehghani, M. Minderer, G. Heigold, S. Gelly, et al. An image is worth 16x16 words: Transformers for image recognition at scale. arXiv preprint arXiv:2010.11929, 2020

2010 arXiv
[16]

Dupont, S

E. Dupont, S. N. Wood, and N. H. Augustin. Spatial+: A novel approach to spatial confounding. Biometrics, 78 0 (4): 0 1279--1290, 2022

2022
[17]

Erickson, J

N. Erickson, J. Mueller, A. Shirkov, H. Zhang, P. Larroy, M. Li, and A. Smola. Autogluon-tabular: Robust and accurate automl for structured data. arXiv preprint arXiv:2003.06505, 2020

2003 arXiv
[18]

Forastiere, E

L. Forastiere, E. M. Airoldi, and F. Mealli. Identification and estimation of treatment and interference effects in observational studies on networks. Journal of the American Statistical Association, 116 0 (534): 0 901--918, 2021

2021
[19]

Frauen, V

D. Frauen, V. Melnychuk, and S. Feuerriegel. Sharp bounds for generalized causal sensitivity analysis. Advances in Neural Information Processing Systems, 2023

2023
[20]

Hamilton, Z

W. Hamilton, Z. Ying, and J. Leskovec. Inductive representation learning on large graphs. Advances in Neural Information Processing Systems, 2017

2017
[21]

E. M. Hanks, E. M. Schliep, M. B. Hooten, and J. A. Hoeting. Restricted spatial regression in practice: geostatistical models, confounding, and robustness under model misspecification. Environmetrics, 26 0 (4): 0 243--254, 2015

2015
[22]

Hatt and S

T. Hatt and S. Feuerriegel. Sequential deconfounding for causal inference with unobserved confounders. In Conference on Causal Learning and Reasoning, 2024

2024
[23]

K. He, X. Zhang, S. Ren, and J. Sun. Deep residual learning for image recognition. In IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2016

2016
[24]

Higgins, L

I. Higgins, L. Matthey, A. Pal, C. Burgess, X. Glorot, M. Botvinick, S. Mohamed, and A. Lerchner. Beta-vae: Learning basic visual concepts with a constrained variational framework. In International Conference on Learning Representations, 2017

2017
[25]

J. L. Hill. Bayesian nonparametric modeling for causal inference. Journal of Computational and Graphical Statistics, 20 0 (1): 0 217--240, 2011

2011
[26]

J. Ho, A. Jain, and P. Abbeel. Denoising diffusion probabilistic models. In Advances in Neural Information Processing Systems, 2020

2020
[27]

J. S. Hodges and B. J. Reich. Adding spatially-correlated errors can mess up the fixed effect you love. The American Statistician, 64 0 (4): 0 325--334, 2010

2010
[28]

M. G. Hudgens and M. E. Halloran. Toward causal inference with interference. Journal of the American Statistical Association, 103 0 (482): 0 832--842, 2008

2008
[29]

G. W. Imbens and D. B. Rubin. Causal inference in statistics, social, and biomedical sciences. Cambridge University Press, 2015

2015
[30]

Kingma, T

D. Kingma, T. Salimans, B. Poole, and J. Ho. Variational diffusion models. In Advances in Neural Information Processing Systems, 2021

2021
[31]

D. P. Kingma and M. Welling. Auto-encoding variational bayes. International Conference on Learning Representations, 2013

2013
[32]

T. Kipf. Semi-supervised classification with graph convolutional networks. arXiv preprint arXiv:1609.02907, 2016

2016 arXiv
[33]

Larsen, S

A. Larsen, S. Yang, B. J. Reich, and A. G. Rappold. A spatial causal analysis of wildland fire-contributed pm2. 5 using numerical model output. The annals of applied statistics, 16 0 (4): 0 2714, 2022

2022
[34]

R. Liaw, E. Liang, R. Nishihara, P. Moritz, J. E. Gonzalez, and I. Stoica. Tune: A research platform for distributed model selection and training, 2018

2018
[35]

Z. Liu, Y. Lin, Y. Cao, H. Hu, Y. Wei, Z. Zhang, S. Lin, and B. Guo. Swin transformer: Hierarchical vision transformer using shifted windows. In IEEE/CVF International Conference on Computer Vision, 2021

2021
[36]

Maal e, C

L. Maal e, C. K. S nderby, S. K. S nderby, and O. Winther. Auxiliary deep generative models. In International Conference on Machine Learning, 2016

2016
[37]

Oktay, J

O. Oktay, J. Schlemper, L. L. Folgoc, M. Lee, M. Heinrich, K. Misawa, K. Mori, S. McDonagh, N. Y. Hammerla, B. Kainz, et al. Attention u - n et: Learning where to look for the pancreas. arXiv preprint arXiv:1804.03999, 2018

2018 arXiv
[38]

Oprescu, J

M. Oprescu, J. Dorn, M. Ghoummaid, A. Jesson, N. Kallus, and U. Shalit. B-learner: Quasi-oracle bounds on heterogeneous causal effects under hidden confounding. In International Conference on Machine Learning, 2023

2023
[39]

Oprescu, D

M. Oprescu, D. K. Park, X. Luo, S. Yoo, and N. Kallus. Gst-unet: Spatiotemporal causal inference with time-varying confounders. arXiv preprint arXiv:2502.05295, 2025

2025
[41]

Papadogeorgou, C

G. Papadogeorgou, C. Choirat, and C. M. Zigler. Adjusting for unmeasured spatial confounding with distance adjusted propensity score matching. Biostatistics, 20 0 (2): 0 256--272, 2019

2019
[42]

Rezende and S

D. Rezende and S. Mohamed. Variational inference with normalizing flows. In International Conference on Machine Learning, 2015

2015
[43]

D. J. Rezende, S. Mohamed, and D. Wierstra. Stochastic backpropagation and approximate inference in deep generative models. In International Conference on Machine Learning, 2014

2014
[44]

D. R. Roberts, V. Bahn, S. Ciuti, M. S. Boyce, J. Elith, G. Guillera-Arroita, S. Hauenstein, J. J. Lahoz-Monfort, B. Schröder, W. Thuiller, D. I. Warton, B. A. Wintle, F. Hartig, and C. F. Dormann. Cross-validation strategies for data with temporal, spatial, hierarchical, or p...

2017
[45]

Ronneberger, P

O. Ronneberger, P. Fischer, and T. Brox. U-net: Convolutional networks for biomedical image segmentation. In Medical Image Computing and Computer-Assisted Intervention -- MICCAI 2015, 2015

2015
[46]

D. B. Rubin. Bayesianly justifiable and relevant frequency calculations for the applied statistician. The Annals of Statistics, pages 1151--1172, 1984

1984
[47]

D. B. Rubin. Causal inference using potential outcomes: Design, modeling, decisions. Journal of the American Statistical Association, 100 0 (469): 0 322--331, 2005

2005
[48]

Rue and L

H. Rue and L. Held. Gaussian Markov random fields: theory and applications. Chapman and Hall/CRC, 2005

2005
[49]

C. Shi, D. M. Blei, and V. Veitch. Adapting neural networks for the estimation of treatment effects. In Advances in Neural Information Processing Systems, 2019

2019
[50]

X. Shi, Z. Chen, H. Wang, D.-Y. Yeung, W.-K. Wong, and W.-c. Woo. Convolutional lstm network: A machine learning approach for precipitation nowcasting. Advances in Neural Information Processing Systems, 2015

2015
[51]

M. E. Sobel. What do randomized studies of housing mobility demonstrate? causal inference in the face of interference. Journal of the American Statistical Association, 101 0 (476): 0 1398--1407, 2006

2006
[52]

K. Sohn, H. Lee, and X. Yan. Learning structured output representation using deep conditional generative models. Advances in Neural Information Processing Systems, 2015

2015
[53]

E. J. Tchetgen Tchetgen, I. R. Fulcher, and I. Shpitser. Auto-g-computation of causal effects on a network. Journal of the American Statistical Association, 116 0 (534): 0 833--844, 2021

2021
[54]

M. Tec, A. Trisovic, M. Audirac, S. Woodward, J. Hu, N. Khoshnevis, and F. Dominici. Spa CE : The spatial confounding environment. In International Conference on Representation Learning, 2024

2024
[55]

Census Bureau

U.S. Census Bureau . 2010 census, 2010

2010
[56]

T. J. VanderWeele, E. J. T. Tchetgen, and M. E. Halloran. Interference and sensitivity analysis. Statistical science: A Review Journal of the Institute of Mathematical Statistics, 29 0 (4): 0 687, 2015

2015
[57]

Veli c kovi \'c , G

P. Veli c kovi \'c , G. Cucurull, A. Casanova, A. Romero, P. Lio, and Y. Bengio. Graph attention networks. arXiv preprint arXiv:1710.10903, 2017

2017 arXiv
[58]

Wang and D

Y. Wang and D. M. Blei. The blessings of multiple causes. Journal of the American Statistical Association, 114 0 (528): 0 1574--1596, 2019

2019
[59]

Zigler, V

C. Zigler, V. Liu, F. Mealli, and L. Forastiere. Bipartite interference and air pollution transport: estimating health effects of power plant interventions. Biostatistics, 26 0 (1): 0 kxae051, 2025

2025
[60]

@esa (Ref

\@ifxundefined[1] #1\@undefined \@firstoftwo \@secondoftwo \@ifnum[1] #1 \@firstoftwo \@secondoftwo \@ifx[1] #1 \@firstoftwo \@secondoftwo [2] @ #1 \@temptokena #2 #1 @ \@temptokena \@ifclassloaded agu2001 natbib The agu2001 class already includes natbib coding, so you should ...
[61]

\@lbibitem[] @bibitem@first@sw\@secondoftwo \@lbibitem[#1]#2 \@extra@b@citeb \@ifundefined br@#2\@extra@b@citeb \@namedef br@#2 \@nameuse br@#2\@extra@b@citeb \@ifundefined b@#2\@extra@b@citeb @num @parse #2 @tmp #1 NAT@b@open@#2 NAT@b@shut@#2 \@ifnum @merge>\@ne @bibitem@firs...
[62]

Hyperlocal super-learned PM 2.5 components across the contiguous US

@open @close @open @close and [1] URL: #1 \@ifundefined chapter * \@mkboth \@ifxundefined @sectionbib * \@mkboth * \@mkboth\@gobbletwo \@ifclassloaded amsart * \@ifclassloaded amsbook * \@ifxundefined @heading @heading NAT@ctr thebibliography [1] @ \@biblabel @NAT@ctr \@bibset...

2023

Pith tools

Reviewed August 4, 2026 · model on record in the stance chip above.

[1] [1]

write newline

" write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 global.max substring 't := if while FUNCTION format.date year duplicate empty "emp...

[2] [2]

S. Ali, O. Faruque, and J. Wang. Estimating direct and indirect causal effects of spatiotemporal interventions in presence of spatial interference. In Joint European Conference on Machine Learning and Knowledge Discovery in Databases, 2024

2024

[3] [3]

Amini, M

H. Amini, M. Danesh-Yazdi, Q. Di, W. Requia, Y. Wei, Y. Abu-Awad, L. Shi, M. Franklin, C.-M. Kang, J. Wolfson, P. James, R. Habre, Q. Zhu, J. Apte, Z. Andersen, I. kloog, F. Dominici, P. Koutrakis, and J. Schwartz. Hyperlocal super-learned PM 2.5 components across the contiguous us, 2022

2022

[4] [4]

J. D. Angrist, G. W. Imbens, and D. B. Rubin. Identification of causal effects using instrumental variables. Journal of the American Statistical Association, 91 0 (434): 0 444--455, 1996

1996

[5] [5]

L. Anselin. Spatial Econometrics: Methods and Models, volume 4. Springer Science & Business Media, 1988

1988

[6] [6]

P. M. Aronow and C. Samii. Estimating average causal effects under general interference, with application to a social network experiment. The Annals of Applied Statistics, 11 0 (4): 0 1912–--1947, 2017

1912

[7] [7]

I. Bica, A. Alaa, and M. Van Der Schaar. Time series deconfounder: Estimating treatment effects over time in the presence of hidden confounders. In International conference on machine learning, 2020

2020

[8] [8]

Behavioral risk factor surveillance system ( BRFSS ), 2010

Centers for Disease Control and Prevention . Behavioral risk factor surveillance system ( BRFSS ), 2010

2010