aiqbench analysis
This research advances causal inference by integrating diffusion models to adapt smoothing operations for high-dimensional data concentrated on lower-dimensional manifolds. By aligning estimators with local outcome geometry, the method achieves intrinsic scaling and improved stability over standard isotropic approaches in complex embedding spaces.
Geometry Adaptive Counterfactual Distribution Learning with Diffusion-Guided Smoothing
arXiv:2605.25811v2 Announce Type: replace-cross Abstract: We study counterfactual distribution learning for high-dimensional outcomes whose laws may concentrate near lower-dimensional structure. Standard isotropic smoothing ignores this geometry, leading to unfavorable scaling and unstable local inference. We propose semiparametrically debiased, diffusion-guided estimators for smoothed counterfactual densities and their scores. These estimators combine causal nuisance adjustment with geometry-adaptive localization driven by a learned diffusion score, yielding second-order nuisance remainders while aligning smoothing with local outcome geometry. We derive asymptotic expansions, integrated risk bounds, and simultaneous inference for smoothed densities and Stein functionals, with extensions to ambient density and score targets under additional approximation conditions. The variance term in the risk bounds is governed by the concentration of the smoothing operator: suitable geometric conditions yield intrinsic rather than ambient scaling, while an explicit drift term quantifies the cost of estimating the geometry. CelebA-based semi-synthetic experiments show faster error decay and improved stability for geometry-adaptive one-step methods, illustrating their applicability to high-dimensional embeddings.
Abstract:We study counterfactual distribution learning for high-dimensional outcomes whose laws may concentrate near lower-dimensional structure. Standard isotropic smoothing ignores this geometry, leading to unfavorable scaling and unstable local inference. We propose semiparametrically debiased, diffusion-guided estimators for smoothed counterfactual densities and their scores. These estimators combine causal nuisance adjustment with geometry-adaptive localization driven by a learned diffusion score, yielding second-order nuisance remainders while aligning smoothing with local outcome geometry. We derive asymptotic expansions, integrated risk bounds, and simultaneous inference for smoothed densities and Stein functionals, with extensions to ambient density and score targets under additional approximation conditions. The variance term in the risk bounds is governed by the concentration of the smoothing operator: suitable geometric conditions yield intrinsic rather than ambient scaling, while an explicit drift term quantifies the cost of estimating the geometry. CelebA-based semi-synthetic experiments show faster error decay and improved stability for geometry-adaptive one-step methods, illustrating their applicability to high-dimensional embeddings.
Methodology (stat.ME); Machine Learning (cs.LG); Machine Learning (stat.ML)
arXiv:2605.25811 [stat.ME]
(or arXiv:2605.25811v2 [stat.ME] for this version)
https://doi.org/10.48550/arXiv.2605.25811 Focus to learn more arXiv-issued DOI via DataCite
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 250 words
Related coverage
MemCorr-DP: Counterfactual Correspondence Conditioning for a Diffusion Policy Guided by a Reference
arXiv · Sep 9, 2026
Our response to Séb Krier on Plan A
LessWrong · Jul 14, 2026
The Machines that Make the Machines
NVIDIA · Oct 7, 2026
RadioTrace: Transmitter-Aware Diffusion for Radio Map Estimation without Deployment-Time Fine-Tuning
arXiv · Jul 24, 2026