Tagging-Augmented Generation: Assisting Language Models in Finding Intricate Knowledge In Long Contexts

TL;DR

This paper introduces Tagging-Augmented Generation (TAG), a lightweight data augmentation method that enhances large language models' ability to handle long contexts in question-answering tasks without complex pre-processing.

Contribution

The authors propose a novel tagging-based augmentation strategy that improves LLM performance on long-context QA benchmarks, avoiding the drawbacks of retrieval and chunking methods.

Findings

Up to 17% performance improvement on 32K token contexts

2.9% gain in complex multi-hop reasoning questions

Effective augmentation without altering document integrity

Abstract

Recent investigations into effective context lengths of modern flagship large language models (LLMs) have revealed major limitations in effective question answering (QA) and reasoning over long and complex contexts for even the largest and most impressive cadre of models. While approaches like retrieval-augmented generation (RAG) and chunk-based re-ranking attempt to mitigate this issue, they are sensitive to chunking, embedding and retrieval strategies and models, and furthermore, rely on extensive pre-processing, knowledge acquisition and indexing steps. In this paper, we propose Tagging-Augmented Generation (TAG), a lightweight data augmentation strategy that boosts LLM performance in long-context scenarios, without degrading and altering the integrity and composition of retrieved documents. We validate our hypothesis by augmenting two challenging and directly relevant question-answering benchmarks -- NoLima and NovelQA -- and show that tagging the context or even just adding tag definitions into QA prompts leads to consistent performance gains over the baseline -- up to 17% for 32K token contexts, and 2.9% in complex reasoning question-answering for multi-hop queries requiring knowledge across a wide span of text. Additional details are available at https://sites.google.com/view/tag-emnlp.