Auto-Instruct: Automatic Instruction Generation and Ranking for Black-Box Language Models

TL;DR

Auto-Instruct automatically generates and ranks high-quality instructions for large language models, improving task performance without manual effort by leveraging LLMs' generative abilities and a trained scoring model.

Contribution

It introduces a novel automatic instruction generation and ranking method that outperforms human-written instructions and generalizes across different LLMs.

Findings

Auto-Instruct surpasses human-written instructions in 118 out-of-domain tasks.

The method leverages LLMs' generative ability for diverse instruction creation.

It demonstrates strong generalizability to other LLMs not used in training.

Abstract

Large language models (LLMs) can perform a wide range of tasks by following natural language instructions, without the necessity of task-specific fine-tuning. Unfortunately, the performance of LLMs is greatly influenced by the quality of these instructions, and manually writing effective instructions for each task is a laborious and subjective process. In this paper, we introduce Auto-Instruct, a novel method to automatically improve the quality of instructions provided to LLMs. Our method leverages the inherent generative ability of LLMs to produce diverse candidate instructions for a given task, and then ranks them using a scoring model trained on a variety of 575 existing NLP tasks. In experiments on 118 out-of-domain tasks, Auto-Instruct surpasses both human-written instructions and existing baselines of LLM-generated instructions. Furthermore, our method exhibits notable generalizability even with other LLMs that are not incorporated into its training process.