Layer by Layer: Uncovering Where Multi-Task Learning Happens in Instruction-Tuned Large Language Models

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Zhao, Zheng, Ziser, Yftah, Cohen, Shay B.
Formato: Preprint
Publicado: 2024
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866929561178996736
author Zhao, Zheng
Ziser, Yftah
Cohen, Shay B.
author_facet Zhao, Zheng
Ziser, Yftah
Cohen, Shay B.
contents Fine-tuning pre-trained large language models (LLMs) on a diverse array of tasks has become a common approach for building models that can solve various natural language processing (NLP) tasks. However, where and to what extent these models retain task-specific knowledge remains largely unexplored. This study investigates the task-specific information encoded in pre-trained LLMs and the effects of instruction tuning on their representations across a diverse set of over 60 NLP tasks. We use a set of matrix analysis tools to examine the differences between the way pre-trained and instruction-tuned LLMs store task-specific information. Our findings reveal that while some tasks are already encoded within the pre-trained LLMs, others greatly benefit from instruction tuning. Additionally, we pinpointed the layers in which the model transitions from high-level general representations to more task-oriented representations. This finding extends our understanding of the governing mechanisms of LLMs and facilitates future research in the fields of parameter-efficient transfer learning and multi-task learning.
format Preprint
id arxiv_https___arxiv_org_abs_2410_20008
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Layer by Layer: Uncovering Where Multi-Task Learning Happens in Instruction-Tuned Large Language Models
Zhao, Zheng
Ziser, Yftah
Cohen, Shay B.
Computation and Language
Machine Learning
Fine-tuning pre-trained large language models (LLMs) on a diverse array of tasks has become a common approach for building models that can solve various natural language processing (NLP) tasks. However, where and to what extent these models retain task-specific knowledge remains largely unexplored. This study investigates the task-specific information encoded in pre-trained LLMs and the effects of instruction tuning on their representations across a diverse set of over 60 NLP tasks. We use a set of matrix analysis tools to examine the differences between the way pre-trained and instruction-tuned LLMs store task-specific information. Our findings reveal that while some tasks are already encoded within the pre-trained LLMs, others greatly benefit from instruction tuning. Additionally, we pinpointed the layers in which the model transitions from high-level general representations to more task-oriented representations. This finding extends our understanding of the governing mechanisms of LLMs and facilitates future research in the fields of parameter-efficient transfer learning and multi-task learning.
title Layer by Layer: Uncovering Where Multi-Task Learning Happens in Instruction-Tuned Large Language Models
topic Computation and Language
Machine Learning
url https://arxiv.org/abs/2410.20008