Tree Prompting: Efficient Task Adaptation without Fine-Tuning

TL;DR

Tree Prompting constructs a decision tree of prompts to adapt language models to new tasks efficiently, improving accuracy without fine-tuning and enabling interpretability.

Contribution

This paper introduces Tree Prompting, a novel prompting method that builds a decision tree of prompts for better task adaptation and interpretability without fine-tuning.

Findings

Outperforms existing prompting methods on classification tasks

Achieves accuracy comparable to fine-tuning for smaller LMs

Enables inspection of model decision-making process

Abstract

Prompting language models (LMs) is the main interface for applying them to new tasks. However, for smaller LMs, prompting provides low accuracy compared to gradient-based finetuning. Tree Prompting is an approach to prompting which builds a decision tree of prompts, linking multiple LM calls together to solve a task. At inference time, each call to the LM is determined by efficiently routing the outcome of the previous call using the tree. Experiments on classification datasets show that Tree Prompting improves accuracy over competing methods and is competitive with fine-tuning. We also show that variants of Tree Prompting allow inspection of a model's decision-making process.