Key points are not available for this paper at this time.
Large Language Models (LLMs) have demonstrated remarkable zero-shot generalization across various language-related tasks, including search engines. However, existing work utilizes the generative ability of LLMs for Information Retrieval (IR) rather than direct passage ranking. The discrepancy between the pre-training objectives of LLMs and the ranking objective poses another challenge. In this paper, we first investigate generative LLMs such as ChatGPT and GPT-4 for relevance ranking in IR. Surprisingly, our experiments reveal that properly instructed LLMs can deliver competitive, even superior results to state-of-the-art supervised methods on popular IR benchmarks. Furthermore, to address concerns about data contamination of LLMs, we collect a new test set called NovelEval, based on the latest knowledge and aiming to verify the model's ability to rank unknown knowledge. Finally, to improve efficiency in real-world applications, we delve into the potential for distilling the ranking capabilities of ChatGPT into small specialized models using a permutation distillation scheme. Our evaluation results turn out that a distilled 440M model outperforms a 3B supervised model on the BEIR benchmark. The code to reproduce our results is available at www.github.com/sunnweiwei/RankGPT.
Building similarity graph...
Analyzing shared references across papers
Loading...
Weiwei Sun
Lingyong Yan
Xinyu Ma
Leiden University
Shandong University
Baidu (China)
Building similarity graph...
Analyzing shared references across papers
Loading...
Sun et al. (Sun,) studied this question.
www.synapsesocial.com/papers/69dcb76ef7297818863592ca — DOI: https://doi.org/10.18653/v1/2023.emnlp-main.923