Manifold-based Verbalizer Space Re-embedding for Tuning-free Prompt-based Classification

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Wang, Haochun, Zhao, Sendong, Liu, Chi, Xi, Nuwa, Cai, Muzhen, Qin, Bing, Liu, Ting
Format: Preprint
Published: 2023
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913213718724608
author Wang, Haochun
Zhao, Sendong
Liu, Chi
Xi, Nuwa
Cai, Muzhen
Qin, Bing
Liu, Ting
author_facet Wang, Haochun
Zhao, Sendong
Liu, Chi
Xi, Nuwa
Cai, Muzhen
Qin, Bing
Liu, Ting
contents Prompt-based classification adapts tasks to a cloze question format utilizing the [MASK] token and the filled tokens are then mapped to labels through pre-defined verbalizers. Recent studies have explored the use of verbalizer embeddings to reduce labor in this process. However, all existing studies require a tuning process for either the pre-trained models or additional trainable embeddings. Meanwhile, the distance between high-dimensional verbalizer embeddings should not be measured by Euclidean distance due to the potential for non-linear manifolds in the representation space. In this study, we propose a tuning-free manifold-based space re-embedding method called Locally Linear Embedding with Intra-class Neighborhood Constraint (LLE-INC) for verbalizer embeddings, which preserves local properties within the same class as guidance for classification. Experimental results indicate that even without tuning any parameters, our LLE-INC is on par with automated verbalizers with parameter tuning. And with the parameter updating, our approach further enhances prompt-based tuning by up to 3.2%. Furthermore, experiments with the LLaMA-7B&13B indicate that LLE-INC is an efficient tuning-free classification approach for the hyper-scale language models.
format Preprint
id arxiv_https___arxiv_org_abs_2309_04174
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Manifold-based Verbalizer Space Re-embedding for Tuning-free Prompt-based Classification
Wang, Haochun
Zhao, Sendong
Liu, Chi
Xi, Nuwa
Cai, Muzhen
Qin, Bing
Liu, Ting
Computation and Language
Artificial Intelligence
Prompt-based classification adapts tasks to a cloze question format utilizing the [MASK] token and the filled tokens are then mapped to labels through pre-defined verbalizers. Recent studies have explored the use of verbalizer embeddings to reduce labor in this process. However, all existing studies require a tuning process for either the pre-trained models or additional trainable embeddings. Meanwhile, the distance between high-dimensional verbalizer embeddings should not be measured by Euclidean distance due to the potential for non-linear manifolds in the representation space. In this study, we propose a tuning-free manifold-based space re-embedding method called Locally Linear Embedding with Intra-class Neighborhood Constraint (LLE-INC) for verbalizer embeddings, which preserves local properties within the same class as guidance for classification. Experimental results indicate that even without tuning any parameters, our LLE-INC is on par with automated verbalizers with parameter tuning. And with the parameter updating, our approach further enhances prompt-based tuning by up to 3.2%. Furthermore, experiments with the LLaMA-7B&13B indicate that LLE-INC is an efficient tuning-free classification approach for the hyper-scale language models.
title Manifold-based Verbalizer Space Re-embedding for Tuning-free Prompt-based Classification
topic Computation and Language
Artificial Intelligence
url https://arxiv.org/abs/2309.04174