The use of relative representations for latent embeddings has shown potential in enabling latent space communication and zero-shot model stitching across a wide range of applications. Nevertheless, relative representations rely on a certain amount of parallel anchors to be given as input, which can be impractical to obtain in certain scenarios. To overcome this limitation, we propose an optimization-based method to discover new parallel anchors from a limited known set (seed). Our approach can be used to find semantic correspondence between different domains, align their relative spaces, and achieve competitive results in several tasks.
@article{arxiv.2303.00721,
title = {Bootstrapping Parallel Anchors for Relative Representations},
author = {Irene Cannistraci and Luca Moschella and Valentino Maiorca and Marco Fumero and Antonio Norelli and Emanuele Rodolà},
journal= {arXiv preprint arXiv:2303.00721},
year = {2023}
}