analiza aiqbench
This paper introduces a novel methodological framework for causal inference by replacing standard cross-world interventions with backtracking counterfactuals to define direct, indirect, and total effects. By constructing alternative exogenous states via conditional distributions, the authors offer a distinct mathematical approach to mediation analysis that could refine how complex AI models interpret treatment pathways.
Backtracking Direct, Indirect and Total Effects
arXiv:2610.28910v1 Announce Type: cross Abstract: Natural direct and indirect effects are defined through cross-world interventional counterfactuals. We propose an alternative construction based on backtracking counterfactuals. Rather than intervening on the treatment or mediator, we use a backtracking conditional distribution to construct an alternative exogenous state satisfying a desired treatment-mediator constraint. This yields the backtracking direct and indirect effects, and also the backtracking total effect. Likewise, we adapt the population intervention effects.
Abstract:Natural direct and indirect effects are defined through cross-world interventional counterfactuals. We propose an alternative construction based on backtracking counterfactuals. Rather than intervening on the treatment or mediator, we use a backtracking conditional distribution to construct an alternative exogenous state satisfying a desired treatment-mediator constraint. This yields the backtracking direct and indirect effects, and also the backtracking total effect. Likewise, we adapt the population intervention effects.
Methodology (stat.ME); Machine Learning (cs.LG); Machine Learning (stat.ML)
arXiv:2610.28910 [stat.ME]
(or arXiv:2610.28910v1 [stat.ME] for this version)
https://doi.org/10.48550/arXiv.2610.28910 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
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.
Text integral sub licența deschisă a publicației · 165 cuvinte
