QQJ: Quantifying Qualitative Judgment for Scalable and Human-Aligned Evaluation of Generative AI

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Veysi, Marjan, Shamsinejadbabaki, Pirooz, Zare, Mohammad, Sabouri, Mohammad
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916020672790528
author Veysi, Marjan
Shamsinejadbabaki, Pirooz
Zare, Mohammad
Sabouri, Mohammad
author_facet Veysi, Marjan
Shamsinejadbabaki, Pirooz
Zare, Mohammad
Sabouri, Mohammad
contents The rapid progress of generative artificial intelligence has exposed fundamental limitations in existing evaluation methodologies, particularly for open-ended, creative, and human-facing tasks. Traditional automatic metrics rely on surface-level statistical similarity and often fail to reflect human perceptions of quality, while purely human evaluation, although reliable, is costly, subjective, and difficult to scale. Recent approaches using large language models as evaluators offer improved scalability but frequently lack explicit grounding in human-defined evaluation principles, leading to bias and inconsistency. In this paper, we introduce Quantifying Qualitative Judgment (QQJ), a scalable and human-centric evaluation framework that explicitly bridges the gap between human judgment and automated assessment. QQJ separates the definition of quality from its execution by anchoring evaluation in expert-designed, multi-dimensional rubrics and calibrating large language model evaluators to align with expert reasoning using a small, high-quality annotation set. This design enables consistent, interpretable, and scalable evaluation across diverse generative tasks and modalities. Extensive experiments on text and image generation demonstrate that QQJ achieves substantially stronger alignment with human judgment than traditional automatic metrics and unconstrained LLM-based evaluators. Moreover, QQJ exhibits improved stability across repeated evaluations and superior diagnostic capability in identifying critical failure modes such as hallucination and intent mismatch. These results indicate that structured qualitative judgment can be operationalized at scale without sacrificing interpretability or human alignment, positioning QQJ as a practical foundation for reliable evaluation of modern generative AI systems.
format Preprint
id arxiv_https___arxiv_org_abs_2605_17382
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle QQJ: Quantifying Qualitative Judgment for Scalable and Human-Aligned Evaluation of Generative AI
Veysi, Marjan
Shamsinejadbabaki, Pirooz
Zare, Mohammad
Sabouri, Mohammad
Artificial Intelligence
Computation and Language
Graphics
The rapid progress of generative artificial intelligence has exposed fundamental limitations in existing evaluation methodologies, particularly for open-ended, creative, and human-facing tasks. Traditional automatic metrics rely on surface-level statistical similarity and often fail to reflect human perceptions of quality, while purely human evaluation, although reliable, is costly, subjective, and difficult to scale. Recent approaches using large language models as evaluators offer improved scalability but frequently lack explicit grounding in human-defined evaluation principles, leading to bias and inconsistency. In this paper, we introduce Quantifying Qualitative Judgment (QQJ), a scalable and human-centric evaluation framework that explicitly bridges the gap between human judgment and automated assessment. QQJ separates the definition of quality from its execution by anchoring evaluation in expert-designed, multi-dimensional rubrics and calibrating large language model evaluators to align with expert reasoning using a small, high-quality annotation set. This design enables consistent, interpretable, and scalable evaluation across diverse generative tasks and modalities. Extensive experiments on text and image generation demonstrate that QQJ achieves substantially stronger alignment with human judgment than traditional automatic metrics and unconstrained LLM-based evaluators. Moreover, QQJ exhibits improved stability across repeated evaluations and superior diagnostic capability in identifying critical failure modes such as hallucination and intent mismatch. These results indicate that structured qualitative judgment can be operationalized at scale without sacrificing interpretability or human alignment, positioning QQJ as a practical foundation for reliable evaluation of modern generative AI systems.
title QQJ: Quantifying Qualitative Judgment for Scalable and Human-Aligned Evaluation of Generative AI
topic Artificial Intelligence
Computation and Language
Graphics
url https://arxiv.org/abs/2605.17382