A Survey on LLM-as-a-Judge
By Jiawei Gu, Xuhui Jiang, Zhichao Shi, Hexiang Tan, Xuehao Zhai, Chengjin Xu, Wei Li, Yinghan Shen, Shengjie Ma, Honghao Liu and 2 more. arXiv.org, 2024
1,583 citations recorded by the source corpus.
Accurate and consistent evaluation is crucial for decision-making across numerous fields, yet it remains a challenging task due to inherent subjectivity, variability, and scale. Large Language Models (LLMs) have achieved remarkable success across diverse domains, leading to the emergence of"LLM-as-a-Judge,"where LLMs are employed as evaluators for complex tasks. With their ability to process diverse data types and provide scalable, cost-effective, and consistent assessments, LLMs present a compelling alternative to traditional expert-driven evaluations. However, ensuring the reliability of LLM-as-a-Judge systems remains a significant challenge that requires careful design and standardization. This paper provides a comprehensive survey of LLM-as-a-Judge, addressing the core question: How can reliable LLM-as-a-Judge systems be built? We explore strategies to enhance reliability, including improving consistency, mitigating biases, and adapting to diverse assessment scenarios. Additionally, we propose methodologies for evaluating the reliability of LLM-as-a-Judge systems, supported by a novel benchmark designed for this purpose. To advance the development and real-world deployment of LLM-as-a-Judge systems, we also discussed practical applications, challenges, and future directions. This survey serves as a foundational reference for researchers and practitioners in this rapidly evolving field.
Subfields: Law, Political Science and International Relations, Strategy and Management
View Paper: https://www.semanticscholar.org/paper/e24424283c02fbe7f641e5b3490d7bb059f8355a
Citations in the corpus
- From Generation to Judgment: Opportunities and Challenges of LLM-as-a-judge (2024, neutral): “2023; Beigi et al., 2024b; Bai et al., 2023a) leads to fine-grained and detailed judgment for various evaluation applications, significantly addressing the limitations of traditional evaluation methods, setting a new standard for what is achievable in NLP evaluation.”
- A framework to assess clinical safety and hallucination rates of LLMs for medical text summarisation (2025, neutral): “‘ LLM-as-a-Judge ’ , which refers to an LLM tasked with evaluating, scoring, or assessing the quality, correctness, or appropriateness of outputs, has recently been described 55 and its applicability has been discussed 56 .”
- HistoRAG: Embedding Historical Methodology in Retrieval-Augmented Generation Through Critical Technical Practice (2026, unknown): “…paradigm, in which instruction-following models evaluate content against explicit criteria, has emerged as a promising post-retrieval strategy (Gu et al., 2024), and Nogueira and Cho (2020) demonstrated that BERT-based passage re-ranking substantially outperforms initial retrieval on standard…”
- A Large-Language-Model Framework for Automated Humanitarian Situation Reporting (2025, unknown): “The prompt used for this classification is reported in Supplementary Material S.0.13 .”
- Agentic Software Issue Resolution with Large Language Models: A Survey (2025, unknown): “LLM-as-a-Judge [56] is another commonly used reranking method [4, 13, 45, 51, 91, 115, 123, 133], which leverages the LLM to identify differences among candidate patches and select the one most aligned with the intent of the issue description.”