Decoding-based Regression

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Song, Xingyou, Bahri, Dara
Formato: Preprint
Publicado: 2025
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866918122324230144
author Song, Xingyou
Bahri, Dara
author_facet Song, Xingyou
Bahri, Dara
contents Language models have recently been shown capable of performing regression wherein numeric predictions are represented as decoded strings. In this work, we provide theoretical grounds for this capability and furthermore investigate the utility of causal sequence decoding models as numeric regression heads given any feature representation. We find that, despite being trained in the usual way - for next-token prediction via cross-entropy loss - decoder-based heads are as performant as standard pointwise heads when benchmarked over standard regression tasks, while being flexible enough to capture smooth numeric distributions, such as in the task of density estimation.
format Preprint
id arxiv_https___arxiv_org_abs_2501_19383
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Decoding-based Regression
Song, Xingyou
Bahri, Dara
Machine Learning
Artificial Intelligence
Computation and Language
Language models have recently been shown capable of performing regression wherein numeric predictions are represented as decoded strings. In this work, we provide theoretical grounds for this capability and furthermore investigate the utility of causal sequence decoding models as numeric regression heads given any feature representation. We find that, despite being trained in the usual way - for next-token prediction via cross-entropy loss - decoder-based heads are as performant as standard pointwise heads when benchmarked over standard regression tasks, while being flexible enough to capture smooth numeric distributions, such as in the task of density estimation.
title Decoding-based Regression
topic Machine Learning
Artificial Intelligence
Computation and Language
url https://arxiv.org/abs/2501.19383