LLM Chain Ensembles for Scalable and Accurate Data Annotation

TL;DR

This paper presents a chain ensemble approach for large language models that improves data annotation accuracy and reduces costs by routing data based on model confidence, outperforming individual models.

Contribution

Introduces an LLM chain ensemble methodology that enhances annotation accuracy and cost-efficiency by sequentially routing data based on uncertainty.

Findings

Outperforms individual LLMs in accuracy.

Achieves significant cost savings.

Effective in large-scale data annotation.

Abstract

The ability of large language models (LLMs) to perform zero-shot classification makes them viable solutions for data annotation in rapidly evolving domains where quality labeled data is often scarce and costly to obtain. However, the large-scale deployment of LLMs can be prohibitively expensive. This paper introduces an LLM chain ensemble methodology that aligns multiple LLMs in a sequence, routing data subsets to subsequent models based on classification uncertainty. This approach leverages the strengths of individual LLMs within a broader system, allowing each model to handle data points where it exhibits the highest confidence, while forwarding more complex cases to potentially more robust models. Our results show that the chain ensemble method often exceeds the performance of the best individual model in the chain and achieves substantial cost savings, making LLM chain ensembles a practical and efficient solution for large-scale data annotation challenges.