You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
This paper surveys research works in the quickly advancing field ofinstruction tuning (IT), a crucial technique to enhance the capabilities andcontrollability of large language models (LLMs). Instruction tuning refers tothe process of further training LLMs on a dataset consisting of\textsc{(instruction, output)} pairs in a supervised fashion, which bridges thegap between the next-word prediction objective of LLMs and the users' objectiveof having LLMs adhere to human instructions. In this work, we make a systematicreview of the literature, including the general methodology of IT, theconstruction of IT datasets, the training of IT models, and applications todifferent modalities, domains and applications, along with an analysis onaspects that influence the outcome of IT (e.g., generation of instructionoutputs, size of the instruction dataset, etc). We also review the potentialpitfalls of IT along with criticism against it, along with efforts pointing outcurrent deficiencies of existing strategies and suggest some avenues forfruitful research.
URL
Affiliations
Abstract
Translation (by gpt-3.5-turbo)
Summary (by gpt-3.5-turbo)
The text was updated successfully, but these errors were encountered: