Bayesian Multi-Task Transfer Learning for Soft Prompt Tuning

Haeju Lee, Minchan Jeong, Se-Young Yun, Kee-Eung Kim

Submission Type: Regular Long Paper

Submission Track: Efficient Methods for NLP

Submission Track 2: Machine Learning for NLP

Keywords: prompt tuning, parameter-efficient fine-tuning, transfer learning, bayesian method

TL;DR: Our method constructs a global posterior distribution over soft prompts from the source tasks, utilizing it as a prior for the target task to leverage the source posterior for guidance.

Abstract:

Prompt tuning, in which prompts are optimized to adapt large-scale pre-trained language models to downstream tasks instead of fine-tuning the full model parameters, has been shown to be particularly effective when the prompts are trained in the multi-task transfer learning setting. These methods generally involve individually training prompts for each source task and then aggregating them to provide the initialization of the prompt for the target task. However, this approach critically
ignores the fact that some of the source tasks could be negatively or positively interfering with each other. We argue that when we extract knowledge from source tasks via training source prompts, we need to consider this correlation among source tasks for better transfer to target tasks. To this end, we propose a Bayesian approach where we work with the posterior distribution of prompts across source tasks. We obtain representative source prompts corresponding to the samples from
the posterior utilizing Stein Variational Gradient Descent, which are then aggregated to constitute the initial target prompt. We show extensive experimental results on the standard benchmark NLP tasks, where our Bayesian multi-task transfer learning approach outperforms the state-of-the-art methods in many settings. Furthermore, our approach requires no auxiliary models other than the prompt itself, achieving high degree of parameter-efficiency.

Submission Number: 5334