Lowering the numerical precision of model parameters and computations is widely adopted to improve the efficiency of retrieval systems. However, when computing relevance scores between the query and documents in low-precision, we observe spurious ties due to the reduced granularity. This introduces high variability in the results based on tie resolution, making the evaluation less reliable. To address this, we propose a more robust retrieval evaluation protocol designed to reduce score variation. It consists of: (1) High-Precision Scoring (HPS), which upcasts the final scoring step to higher precision to resolve tied candidates with minimal computational cost; and (2) Tie-aware Retrieval Metrics (TRM), which report expected scores, range, and bias to quantify order uncertainty of tied candidates. Our experiments test multiple models with three scoring functions on two retrieval datasets to demonstrate that HPS dramatically reduces tie-induced instability, and TRM accurately recovers expected metric values. This combination enables a more consistent and reliable evaluation system for lower-precision retrievals.

翻译：降低模型参数和计算数值精度被广泛采用以提高检索系统的效率。然而，在低精度下计算查询与文档之间的相关性分数时，我们观察到因粒度降低导致的虚假并列现象。这引入了基于并列消解结果的高变异性，使评估的可靠性下降。为了解决这一问题，我们提出了一种更为稳健的检索评估协议，旨在降低分数变异性。该协议包括：（1）高精度评分（High-Precision Scoring, HPS），它将最终评分步骤提升至更高精度，以最小计算成本消解候选并列项；以及（2）并列感知检索指标（Tie-aware Retrieval Metrics, TRM），它报告期望分数、范围和偏置，以量化并列候选的顺序不确定性。我们的实验在两个检索数据集上使用三种评分函数测试了多个模型，结果表明HPS显著降低了并列引发的不稳定性，且TRM能够准确恢复期望指标值。这一组合为低精度检索提供了更一致、更可靠的评估体系。