Jun 2, 2025
Approximating Human Preferences Using a Multi-Judge Learned System
Eitan Sprejer, Fernando Avalos, José Pedro Brito de Azevedo Faustino, Augusto Mariano Bernardi
Summary
In this work, we introduced a learned approach to aggregating multi-judge scores: using a GAM and a simple MLP as an alternative to traditional, non-learned methods like averaging. Our models outperform the naive baseline in predicting simulated human preferences, demonstrating that learned aggregation can better capture complex evaluative signals.
Cite this work:
@misc {
title={
@misc {
},
author={
Eitan Sprejer, Fernando Avalos, José Pedro Brito de Azevedo Faustino, Augusto Mariano Bernardi
},
date={
6/2/25
},
organization={Apart Research},
note={Research submission to the research sprint hosted by Apart.},
howpublished={https://apartresearch.com}
}