RE-AdaptIR：通過反向工程適應改進信息檢索

RE-AdaptIR: Improving Information Retrieval through Reverse Engineered Adaptation

June 20, 2024

作者: William Fleshman, Benjamin Van Durme

cs.AI

摘要

針對文本檢索進行微調的大型語言模型（LLMs）已在多個信息檢索（IR）基準測試中展示出最先進的結果。然而，為了改進這些模型，監督式訓練需要大量標記的示例，這些通常難以獲得或成本高昂。在這項工作中，我們探索了將逆向工程適應（RE-AdaptIR）擴展到信息檢索上下文的有效性。我們使用RE-AdaptIR僅使用未標記數據來改進基於LLM的IR模型。我們展示了在訓練領域以及模型從未見過查詢的零-shot領域中的性能改進。我們分析了各種微調情境中的性能變化，並提供對從業者立即有用的研究結果。

English

Large language models (LLMs) fine-tuned for text-retrieval have demonstrated state-of-the-art results across several information retrieval (IR) benchmarks. However, supervised training for improving these models requires numerous labeled examples, which are generally unavailable or expensive to acquire. In this work, we explore the effectiveness of extending reverse engineered adaptation to the context of information retrieval (RE-AdaptIR). We use RE-AdaptIR to improve LLM-based IR models using only unlabeled data. We demonstrate improved performance both in training domains as well as zero-shot in domains where the models have seen no queries. We analyze performance changes in various fine-tuning scenarios and offer findings of immediate use to practitioners.