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

TL;DR

This paper introduces a tuning-free, manifold-based re-embedding method called LLE-INC for verbalizer embeddings in prompt-based classification, effectively preserving local class properties and matching tuned methods' performance.

Contribution

Proposes LLE-INC, a novel tuning-free manifold re-embedding technique for verbalizer embeddings that maintains local class structures in high-dimensional spaces.

Findings

LLE-INC matches the performance of tuned verbalizer methods.

LLE-INC improves prompt tuning accuracy by up to 3.2%.

Effective for large-scale language models like LLaMA-7B and 13B.

Abstract

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.