Saved in:
Bibliographic Details
Main Authors: Belanec, Robert, Ostermann, Simon, Srba, Ivan, Bielikova, Maria
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2408.01119
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911676266184704
author Belanec, Robert
Ostermann, Simon
Srba, Ivan
Bielikova, Maria
author_facet Belanec, Robert
Ostermann, Simon
Srba, Ivan
Bielikova, Maria
contents Prompt tuning is an efficient solution for training large language models (LLMs). However, current soft-prompt-based methods often sacrifice multi-task modularity, requiring the training process to be fully or partially repeated for each newly added task. While recent work on task vectors applied arithmetic operations on full model weights to achieve the desired multi-task performance, a similar approach for soft-prompts is still missing. To this end, we introduce Task Prompt Vectors, created by element-wise difference between weights of tuned soft-prompts and their random initialization. Experimental results on 12 NLU datasets show that task prompt vectors can be used in low-resource settings to effectively initialize prompt tuning on similar tasks. In addition, we show that task prompt vectors are independent of the random initialization of prompt tuning on 2 different language model architectures. This allows prompt arithmetics with the pre-trained vectors from different tasks. In this way, we provide a competitive alternative to state-of-the-art baselines by arithmetic addition of task prompt vectors from multiple tasks.
format Preprint
id arxiv_https___arxiv_org_abs_2408_01119
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Task Prompt Vectors: Effective Initialization through Multi-Task Soft-Prompt Transfer
Belanec, Robert
Ostermann, Simon
Srba, Ivan
Bielikova, Maria
Computation and Language
Prompt tuning is an efficient solution for training large language models (LLMs). However, current soft-prompt-based methods often sacrifice multi-task modularity, requiring the training process to be fully or partially repeated for each newly added task. While recent work on task vectors applied arithmetic operations on full model weights to achieve the desired multi-task performance, a similar approach for soft-prompts is still missing. To this end, we introduce Task Prompt Vectors, created by element-wise difference between weights of tuned soft-prompts and their random initialization. Experimental results on 12 NLU datasets show that task prompt vectors can be used in low-resource settings to effectively initialize prompt tuning on similar tasks. In addition, we show that task prompt vectors are independent of the random initialization of prompt tuning on 2 different language model architectures. This allows prompt arithmetics with the pre-trained vectors from different tasks. In this way, we provide a competitive alternative to state-of-the-art baselines by arithmetic addition of task prompt vectors from multiple tasks.
title Task Prompt Vectors: Effective Initialization through Multi-Task Soft-Prompt Transfer
topic Computation and Language
url https://arxiv.org/abs/2408.01119