# Learning plug-in surrogate endpoints for randomized experiments

arXiv: https://arxiv.org/abs/2605.12051 · HTML: https://mlsanity.com/paper/2605.12051/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 4 of its 41 references that are in the map.
- First posted on arXiv 12 May 2026.

## Builds on (most cited first, 4 of 4)

- [PyTorch: An Imperative Style, High-Performance Deep Learning Library](https://mlsanity.com/paper/1912.01703/) (2019)
- [Normalizing Flows for Probabilistic Modeling and Inference](https://mlsanity.com/paper/1912.02762/) (2019)
- [Choosing a Proxy Metric from Past Experiments](https://mlsanity.com/paper/2309.07893/) (2023)
- [Evaluating the Surrogate Index as a Decision-Making Tool Using 200 A/B Tests at Netflix](https://mlsanity.com/paper/2311.11922/) (2023)

## Built on by (most cited first, 1 of 1)

- [Symmetries and Causality: Causal Effect Identification Beyond IID Data](https://mlsanity.com/paper/2609.03697/) (2026)

## Abstract (arXiv)

Surrogate endpoints are used in place of long-term outcomes in randomized experiments when observing the real outcome for a large enough cohort is prohibitively expensive or impractical. A short-term surrogate is good if the result of an experiment using the surrogate is predictive of the result of a hypothetical study using the real outcome. Much attention has been paid to formalizing this property in causal terms, but most criteria are unidentifiable and cannot be turned into practical algorithms for learning surrogate endpoints from data. To address this, we study plug-in composite surrogates, functions of post-treatment variables that may be substituted directly for the primary outcome in a randomized experiment. We propose two methods for learning plug-in surrogates that maximize effect predictiveness, and characterize the possibility of finding endpoints that yield unbiased effect estimates in representative scenarios. Finally, in both synthetic experiments with known effects and in data from a real-world experiment, we find that our method, based on directly modeling the surrogate effect, returns plug-in endpoints more predictive of the primary effect than established methods.