paper-with-me

홈 › Papers

Agreement Metrics for LLM-as-Judge Evaluation: What to Report and Why

2026-05-25 · Delip Rao, Chris Callison-Burch arxiv

Validating an LLM judge against human annotations usually means reporting several agreement statistics: accuracy, precision, recall, $F_1$, Cohen's $κ$, and one or more rank correlations. A survey of 24 recent LLM-as-judge papers finds metric choice entangled with the judgment scale, tie handling, invalid outputs, and abstention handling, and those choices rarely stated. For binary criteria -- the common case in rubric-based evaluation, where each criterion is graded MET or UNMET -- most of the reported numbers are redundant: Pearson's $r$, Spearman's $ρ$, Kendall's $τ_b$, the phi coefficient $φ$, and the Matthews Correlation Coefficient all reduce to a single number on non-degenerate binary data, so reporting several of them only creates an illusion of corroborating evidence. Cohen's $κ$ is the one agreement coefficient that adds information: it shares $φ$'s numerator but normalizes differently, and the gap between them measures how far the judge's positive-label rate has drifted from the human's. We then trace what changes when a judge may abstain with a CANNOT_ASSESS verdict: the three common ways of handling abstentions are not interchangeable preprocessing choices but answer different questions, and they break the binary equivalences. The same equivalences reappear, up to a negligible finite-sample correction, for multi-judge ensembles scored with Fleiss' $κ$ or Krippendorff's $α$. We close with a reporting checklist that names the judgment scale, the abstention and tie handling mode, coverage, the confusion matrix, and the aggregation level alongside any scalar agreement coefficient.

📄 PDF Abstract BibTeX arXiv:2606.00093

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

Who Grades the Grader? Co-Evolving Evaluation Metrics and Skills for Self-Improving LLM Agents

2026-07-14 · Xing Zhang, Guanghui Wang, Yanwei Cui, Ziyuan Li 외 arxiv

Self-evolving agent systems create, revise, and retire their own skills, but every such loop assumes a reliable evaluation metric already exists. In many real applications none does. We show the metric itself can be the …

Code Generation

Blinded Radiologist and LLM-Based Evaluation of LLM-Generated Japanese Translations of Chest CT Reports: Comparative Study

2026-04-02 · Yosuke Yamagishi, Atsushi Takamatsu, Yasunori Hamaguchi, Tomohiro Kikuchi 외 arxiv

Background: Accurate translation of radiology reports is important for multilingual research, clinical communication, and radiology education, but the validity of LLM-based evaluation remains unclear. Objective: To evalu…

VERT: Reliable LLM Judges for Radiology Report Evaluation

2026-04-03 · Federica Bologna, Jean-Philippe Corbeil, Matthew Wilkens, Asma Ben Abacha arxiv

Current literature on radiology report evaluation has focused primarily on designing LLM-based metrics and fine-tuning small models for chest X-rays. However, it remains unclear whether these approaches are robust when a…

parameter-efficient fine-tuning

Trust or Escalate: LLM Judges with Provable Guarantees for Human Agreement

2024-07-25 · JaeHun Jung, Faeze Brahman, Yejin Choi

We present a principled approach to provide LLM-based evaluation with a rigorous guarantee of human agreement. We first propose that a reliable evaluation method should not uncritically rely on model preferences for pair…

Chatbot

Judging the Judges: Evaluating Alignment and Vulnerabilities in LLMs-as-Judges

2024-06-18 · Aman Singh Thakur, Kartik Choudhary, Venkat Srinik Ramayapally, Sankaran Vaidyanathan 외

Offering a promising solution to the scalability challenges associated with human evaluation, the LLM-as-a-judge paradigm is rapidly gaining traction as an approach to evaluating large language models (LLMs). However, th…

TriviaQA