DynaMaR: Dynamic Prompt with Mask Token Representation

TL;DR

DynaMaR introduces a dynamic prompt method with mask token representation that enhances fine-tuning of large language models, reducing overfitting and manual effort, leading to significant performance improvements in various tasks.

Contribution

The paper presents DynaMaR, a novel dynamic prompt approach that addresses overfitting and manual prompt design issues in prompt-based fine-tuning.

Findings

10% average improvement in few-shot settings

3.7% improvement in data-rich settings

Effective across four e-commerce applications

Abstract

Recent research has shown that large language models pretrained using unsupervised approaches can achieve significant performance improvement on many downstream tasks. Typically when adapting these language models to downstream tasks, like a classification or regression task, we employ a fine-tuning paradigm in which the sentence representation from the language model is input to a task-specific head; the model is then fine-tuned end-to-end. However, with the emergence of models like GPT-3, prompt-based fine-tuning has been proven to be a successful approach for few-shot tasks. Inspired by this work, we study discrete prompt technologies in practice. There are two issues that arise with the standard prompt approach. First, it can overfit on the prompt template. Second, it requires manual effort to formulate the downstream task as a language model problem. In this paper, we propose an improvement to prompt-based fine-tuning that addresses these two issues. We refer to our approach as DynaMaR -- Dynamic Prompt with Mask Token Representation. Results show that DynaMaR can achieve an average improvement of 10% in few-shot settings and improvement of 3.7% in data-rich settings over the standard fine-tuning approach on four e-commerce applications.