Rank-without-GPT: Building GPT-Independent Listwise Rerankers on Open-Source Large Language Models
This addresses reproducibility and generalization issues in LLM-based reranking for the information retrieval community, though it is incremental as it adapts existing methods to open-source models.
The paper tackled the problem of building listwise rerankers without dependency on GPT models, achieving a 13% improvement over GPT-3.5-based rerankers and 97% effectiveness compared to GPT-4-based ones in passage retrieval experiments.
Listwise rerankers based on large language models (LLM) are the zero-shot state-of-the-art. However, current works in this direction all depend on the GPT models, making it a single point of failure in scientific reproducibility. Moreover, it raises the concern that the current research findings only hold for GPT models but not LLM in general. In this work, we lift this pre-condition and build for the first time effective listwise rerankers without any form of dependency on GPT. Our passage retrieval experiments show that our best list se reranker surpasses the listwise rerankers based on GPT-3.5 by 13% and achieves 97% effectiveness of the ones built on GPT-4. Our results also show that the existing training datasets, which were expressly constructed for pointwise ranking, are insufficient for building such listwise rerankers. Instead, high-quality listwise ranking data is required and crucial, calling for further work on building human-annotated listwise data resources.