Tree Prompting: Efficient Task Adaptation without Fine-Tuning

Chandan Singh, John Xavier Morris, Alexander M Rush, Jianfeng Gao, Yuntian Deng

Submission Type: Regular Long Paper

Submission Track: Machine Learning for NLP

Submission Track 2: NLP Applications

Keywords: Decision tree, large language model, chain prompting, prompt engineering

TL;DR: A decision tree of prompts can be competitive with fine-tuning

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 fine-tuning. Tree Prompting is an approach to prompting which builds a decision tree of prompts, linking multiple prompt-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.

Submission Number: 2058