Promptagator: Few-shot Dense Retrieval From 8 Examples

TL;DR

This paper introduces Promptagator, a few-shot dense retrieval method using large language models to generate task-specific queries from minimal examples, outperforming traditional models trained on large datasets.

Contribution

The paper presents a novel approach leveraging LLMs for task-specific query generation in few-shot retrieval, eliminating the need for large training datasets.

Findings

Promptagator outperforms models trained on MS MARCO with only 8 examples.

Generated data improves re-ranker performance by 5.0 nDCG points.

Query generation surpasses previous methods in few-shot retrieval settings.

Abstract

Much recent research on information retrieval has focused on how to transfer from one task (typically with abundant supervised data) to various other tasks where supervision is limited, with the implicit assumption that it is possible to generalize from one task to all the rest. However, this overlooks the fact that there are many diverse and unique retrieval tasks, each targeting different search intents, queries, and search domains. In this paper, we suggest to work on Few-shot Dense Retrieval, a setting where each task comes with a short description and a few examples. To amplify the power of a few examples, we propose Prompt-base Query Generation for Retriever (Promptagator), which leverages large language models (LLM) as a few-shot query generator, and creates task-specific retrievers based on the generated data. Powered by LLM's generalization ability, Promptagator makes it possible to create task-specific end-to-end retrievers solely based on a few examples {without} using Natural Questions or MS MARCO to train %question generators or dual encoders. Surprisingly, LLM prompting with no more than 8 examples allows dual encoders to outperform heavily engineered models trained on MS MARCO like ColBERT v2 by more than 1.2 nDCG on average on 11 retrieval sets. Further training standard-size re-rankers using the same generated data yields another 5.0 point nDCG improvement. Our studies determine that query generation can be far more effective than previously observed, especially when a small amount of task-specific knowledge is given.