Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Yang, Kisu, Jang, Yoonna, Jang, Hwanseok, Choi, Kenneth, Augenstein, Isabelle, Lim, Heuiseok
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:https://arxiv.org/abs/2508.03306
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866918440061632512
author Yang, Kisu
Jang, Yoonna
Jang, Hwanseok
Choi, Kenneth
Augenstein, Isabelle
Lim, Heuiseok
author_facet Yang, Kisu
Jang, Yoonna
Jang, Hwanseok
Choi, Kenneth
Augenstein, Isabelle
Lim, Heuiseok
contents 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.
format Preprint
id arxiv_https___arxiv_org_abs_2508_03306
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Reliable Evaluation Protocol for Low-Precision Retrieval
Yang, Kisu
Jang, Yoonna
Jang, Hwanseok
Choi, Kenneth
Augenstein, Isabelle
Lim, Heuiseok
Information Retrieval
Artificial Intelligence
Computation and Language
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.
title Reliable Evaluation Protocol for Low-Precision Retrieval
topic Information Retrieval
Artificial Intelligence
Computation and Language
url https://arxiv.org/abs/2508.03306