Global Prompt Cell: A Portable Control Module for Effective Prompt Tuning

TL;DR

This paper introduces the Global Prompt Cell, a control module that enhances prompt tuning by preserving prompt information across all encoder layers, leading to significant performance improvements.

Contribution

The paper proposes the Global Prompt Cell, a novel portable control module that improves prompt tuning effectiveness by maintaining prompt information throughout model layers.

Findings

Achieved 5.8% improvement on SuperGLUE datasets

Demonstrated effectiveness of GPC in prompt tuning

Enhanced prompt information preservation across layers

Abstract

As a novel approach to tuning pre-trained models, prompt tuning involves freezing the parameters in downstream tasks while inserting trainable embeddings into inputs in the first layer. However, previous methods have mainly focused on the initialization of prompt embeddings. The strategy of training and utilizing prompt embeddings in a reasonable way has become a limiting factor in the effectiveness of prompt tuning. To address this issue, we introduce the Global Prompt Cell (GPC), a portable control module for prompt tuning that selectively preserves prompt information across all encoder layers. Our experimental results demonstrate a 5.8% improvement on SuperGLUE datasets compared to vanilla prompt tuning.