AcuRank: Uncertainty-Aware Adaptive Computation for Listwise Reranking

TL;DR

AcuRank introduces an adaptive reranking method that dynamically allocates computational resources based on uncertainty estimates, improving efficiency and accuracy in listwise reranking with large language models.

Contribution

It presents a novel uncertainty-aware framework using Bayesian TrueSkill for adaptive computation in listwise reranking, enhancing efficiency and scalability.

Findings

Outperforms fixed-computation baselines on TREC-DL and BEIR benchmarks.

Achieves better accuracy-efficiency trade-offs.

Scales more effectively with compute resources.

Abstract

Listwise reranking with large language models (LLMs) enhances top-ranked results in retrieval-based applications. Due to the limit in context size and high inference cost of long context, reranking is typically performed over a fixed size of small subsets, with the final ranking aggregated from these partial results. This fixed computation disregards query difficulty and document distribution, leading to inefficiencies. We propose AcuRank, an adaptive reranking framework that dynamically adjusts both the amount and target of computation based on uncertainty estimates over document relevance. Using a Bayesian TrueSkill model, we iteratively refine relevance estimates until reaching sufficient confidence levels, and our explicit modeling of ranking uncertainty enables principled control over reranking behavior and avoids unnecessary updates to confident predictions. Results on the TREC-DL and BEIR benchmarks show that our method consistently achieves a superior accuracy-efficiency trade-off and scales better with compute than fixed-computation baselines. These results highlight the effectiveness and generalizability of our method across diverse retrieval tasks and LLM-based reranking models.