MultEval: Supporting Collaborative Alignment for LLM-as-a-Judge Evaluation Criteria

TL;DR

MultEval is a system designed to facilitate collaborative creation and refinement of evaluation criteria for LLM-as-a-judge systems, addressing challenges of shared understanding and stakeholder alignment.

Contribution

The paper introduces MultEval, a novel tool supporting multi-stakeholder collaboration in defining and refining evaluation criteria for language model assessments.

Findings

Stakeholders face difficulties in aligning values and translating judgments into criteria.

MultEval enables consensus-building and transparency in criteria development.

Domain experts successfully used MultEval to collaboratively develop evaluation criteria.

Abstract

LLM-as-a-judge approaches have emerged as a scalable solution for evaluating model behaviors, yet they rely on evaluation criteria often created by a single individual, embedding that person's assumptions, priorities, and interpretive lens. In practice, defining such criteria is a collaborative and contested process involving multiple stakeholders with different values, interpretations, and priorities; an aspect largely unsupported by existing tools. To examine this problem in depth, we present a formative study examining how stakeholders collaboratively create, negotiate, and refine evaluation criteria for LLM-as-a-judge systems. Our findings reveal challenges in human oversight, including difficulties in establishing shared understanding, aligning values across stakeholders with different expertise and priorities, and translating nuanced human judgments into criteria that are interpretable and actionable for LLM judges. Based on these insights, we developed MultEval, a system that supports collaborative criteria by enabling multiple evaluators to surface and diagnose disagreements using consensus-building theory, iteratively revise criteria with attached examples and proposal history, and maintain transparency over how judgments are encoded into an automated evaluator. We further report a case study in which a team of domain experts used MultEval to collaboratively author criteria, illustrating how coordination and collaborative consensus-making shape criteria evolution.