Latent Space Translation via Semantic Alignment

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Maiorca, Valentino, Moschella, Luca, Norelli, Antonio, Fumero, Marco, Locatello, Francesco, Rodolà, Emanuele
Format: Preprint
Published: 2023
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911774563893248
author Maiorca, Valentino
Moschella, Luca
Norelli, Antonio
Fumero, Marco
Locatello, Francesco
Rodolà, Emanuele
author_facet Maiorca, Valentino
Moschella, Luca
Norelli, Antonio
Fumero, Marco
Locatello, Francesco
Rodolà, Emanuele
contents While different neural models often exhibit latent spaces that are alike when exposed to semantically related data, this intrinsic similarity is not always immediately discernible. Towards a better understanding of this phenomenon, our work shows how representations learned from these neural modules can be translated between different pre-trained networks via simpler transformations than previously thought. An advantage of this approach is the ability to estimate these transformations using standard, well-understood algebraic procedures that have closed-form solutions. Our method directly estimates a transformation between two given latent spaces, thereby enabling effective stitching of encoders and decoders without additional training. We extensively validate the adaptability of this translation procedure in different experimental settings: across various trainings, domains, architectures (e.g., ResNet, CNN, ViT), and in multiple downstream tasks (classification, reconstruction). Notably, we show how it is possible to zero-shot stitch text encoders and vision decoders, or vice-versa, yielding surprisingly good classification performance in this multimodal setting.
format Preprint
id arxiv_https___arxiv_org_abs_2311_00664
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Latent Space Translation via Semantic Alignment
Maiorca, Valentino
Moschella, Luca
Norelli, Antonio
Fumero, Marco
Locatello, Francesco
Rodolà, Emanuele
Machine Learning
While different neural models often exhibit latent spaces that are alike when exposed to semantically related data, this intrinsic similarity is not always immediately discernible. Towards a better understanding of this phenomenon, our work shows how representations learned from these neural modules can be translated between different pre-trained networks via simpler transformations than previously thought. An advantage of this approach is the ability to estimate these transformations using standard, well-understood algebraic procedures that have closed-form solutions. Our method directly estimates a transformation between two given latent spaces, thereby enabling effective stitching of encoders and decoders without additional training. We extensively validate the adaptability of this translation procedure in different experimental settings: across various trainings, domains, architectures (e.g., ResNet, CNN, ViT), and in multiple downstream tasks (classification, reconstruction). Notably, we show how it is possible to zero-shot stitch text encoders and vision decoders, or vice-versa, yielding surprisingly good classification performance in this multimodal setting.
title Latent Space Translation via Semantic Alignment
topic Machine Learning
url https://arxiv.org/abs/2311.00664