@inproceedings{luan-etal-2026-evaluation,
title = "The Evaluation Scaling Law: Toward More Reliable {LLM} Assessment via Multi-Prompt Diversity",
author = "Luan, Hongzhi and
Huan, Zhaoxin and
Shi, Linfeng and
Zheng, Longfei and
Zhang, Xiaolu and
Zhou, Jun",
editor = "Lee, Hung-Yi and
Dev, Sunipa and
Li, Jessy and
Elliott, Desmond",
booktitle = "Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.emnlp-main.62/",
pages = "1419--1433",
ISBN = "979-8-89176-459-0",
abstract = "Large Language Model (LLM) evaluation is frequently challenged by prompt sensitivity, which limits the reliability of benchmarks relying on a single template. To address this, we systematically investigate a multi-prompt evaluation framework that aggregates performance across diverse templates, leveraging a Quality-Diversity (QD) prompt selection mechanism to ensure both variety and fidelity. Through extensive empirical analysis, we uncover a novel Evaluation Scaling Law: evaluation reliability scales consistently with template diversity. Specifically, our experiments demonstrate that evaluation variance diminishes as the scale of prompts increases, while ranking monotonicity exhibits a log-linear improvement relative to the number of templates. By decoupling a model{'}s intrinsic capability from template-specific fluctuations, our framework facilitates more reliable assessments of LLMs, thereby providing a solid foundation for benchmarking that enables more effective monitoring of training dynamics and rigorous comparisons among state-of-the-art models."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="luan-etal-2026-evaluation">
<titleInfo>
<title>The Evaluation Scaling Law: Toward More Reliable LLM Assessment via Multi-Prompt Diversity</title>
</titleInfo>
<name type="personal">
<namePart type="given">Hongzhi</namePart>
<namePart type="family">Luan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zhaoxin</namePart>
<namePart type="family">Huan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Linfeng</namePart>
<namePart type="family">Shi</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Longfei</namePart>
<namePart type="family">Zheng</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Xiaolu</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jun</namePart>
<namePart type="family">Zhou</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-10</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing</title>
</titleInfo>
<name type="personal">
<namePart type="given">Hung-Yi</namePart>
<namePart type="family">Lee</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sunipa</namePart>
<namePart type="family">Dev</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jessy</namePart>
<namePart type="family">Li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Desmond</namePart>
<namePart type="family">Elliott</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Budapest, Hungary</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-459-0</identifier>
</relatedItem>
<abstract>Large Language Model (LLM) evaluation is frequently challenged by prompt sensitivity, which limits the reliability of benchmarks relying on a single template. To address this, we systematically investigate a multi-prompt evaluation framework that aggregates performance across diverse templates, leveraging a Quality-Diversity (QD) prompt selection mechanism to ensure both variety and fidelity. Through extensive empirical analysis, we uncover a novel Evaluation Scaling Law: evaluation reliability scales consistently with template diversity. Specifically, our experiments demonstrate that evaluation variance diminishes as the scale of prompts increases, while ranking monotonicity exhibits a log-linear improvement relative to the number of templates. By decoupling a model’s intrinsic capability from template-specific fluctuations, our framework facilitates more reliable assessments of LLMs, thereby providing a solid foundation for benchmarking that enables more effective monitoring of training dynamics and rigorous comparisons among state-of-the-art models.</abstract>
<identifier type="citekey">luan-etal-2026-evaluation</identifier>
<location>
<url>https://aclanthology.org/2026.emnlp-main.62/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>1419</start>
<end>1433</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T The Evaluation Scaling Law: Toward More Reliable LLM Assessment via Multi-Prompt Diversity
%A Luan, Hongzhi
%A Huan, Zhaoxin
%A Shi, Linfeng
%A Zheng, Longfei
%A Zhang, Xiaolu
%A Zhou, Jun
%Y Lee, Hung-Yi
%Y Dev, Sunipa
%Y Li, Jessy
%Y Elliott, Desmond
%S Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-459-0
%F luan-etal-2026-evaluation
%X Large Language Model (LLM) evaluation is frequently challenged by prompt sensitivity, which limits the reliability of benchmarks relying on a single template. To address this, we systematically investigate a multi-prompt evaluation framework that aggregates performance across diverse templates, leveraging a Quality-Diversity (QD) prompt selection mechanism to ensure both variety and fidelity. Through extensive empirical analysis, we uncover a novel Evaluation Scaling Law: evaluation reliability scales consistently with template diversity. Specifically, our experiments demonstrate that evaluation variance diminishes as the scale of prompts increases, while ranking monotonicity exhibits a log-linear improvement relative to the number of templates. By decoupling a model’s intrinsic capability from template-specific fluctuations, our framework facilitates more reliable assessments of LLMs, thereby providing a solid foundation for benchmarking that enables more effective monitoring of training dynamics and rigorous comparisons among state-of-the-art models.
%U https://aclanthology.org/2026.emnlp-main.62/
%P 1419-1433
Markdown (Informal)
[The Evaluation Scaling Law: Toward More Reliable LLM Assessment via Multi-Prompt Diversity](https://aclanthology.org/2026.emnlp-main.62/) (Luan et al., EMNLP 2026)
ACL