mlsanityData to 7 Oct 2026

An Empirical Study of LLM-as-a-Judge: How Design Choices Impact Evaluation Reliability

arXiv 2506.13639 ↗First posted 16 Jun 2025cs.CLcited by 17

Not on the map. Why?

Abstract (arXiv)

As large language models (LLMs) continue to advance, reliable evaluation methods are essential particularly for open-ended, instruction-following tasks. LLM-as-a-Judge enables automatic evaluation using LLMs as evaluators, but its reliability remains uncertain. In this work, we analyze key factors affecting its trustworthiness, focusing on alignment with human judgments and evaluation consistency. Using BIGGENBench and EvalBiasBench, we study the effects of evaluation design, decoding strategies, and Chain-of-Tought (CoT) reasoning in evaluation. Our results show that evaluation criteria are critical for reliability, non-deterministic sampling improves alignment with human preferences over deterministic evaluation, and CoT reasoning offers minimal gains when clear evaluation criteria are present.

Key facts. Cited by 17 papers in the map (data through 7 Oct 2026). Builds on 22 of its 43 references that are in the map. First posted on arXiv 16 Jun 2025.

Builds on

22 of 43 references.

References · 22

  1. 2024The Llama 3 Herd of Modelscited by 14,192
  2. 2023Judging LLM-as-a-Judge with MT-Bench and Chatbot Arenacited by 8,393
  3. 2019BERTScore: Evaluating Text Generation with BERTcited by 5,462
  4. 2023Self-Refine: Iterative Refinement with Self-Feedbackcited by 3,317
  5. 2014CIDEr: Consensus-based Image Description Evaluationcited by 2,027
  6. 2024A Survey on LLM-as-a-Judgecited by 1,384
  7. 2020COMET: A Neural Framework for MT Evaluationcited by 1,024
  8. 2024Length-Controlled AlpacaEval: A Simple Way to Debias Automatic Evaluatorscited by 865
  9. 2024From Crowdsourced Data to High-Quality Benchmarks: Arena-Hard and BenchBuilder Pipelinecited by 484
  10. 2024LLMs-as-Judges: A Comprehensive Survey on LLM-based Evaluation Methodscited by 451
  11. 2024Justice or Prejudice? Quantifying Biases in LLM-as-a-Judgecited by 354
  12. 2024Humans or LLMs as the Judge? A Study on Judgement Biasescited by 314
10 more
  1. 2023Evaluating Large Language Models at Evaluating Instruction Followingcited by 290
  2. 2024When Can LLMs Actually Correct Their Own Mistakes? A Critical Survey of Self-Correction of LLMscited by 230
  3. 2023Benchmarking Foundation Models with Language-Model-as-an-Examinercited by 172
  4. 2024Aligning with Human Judgement: The Role of Pairwise Preference in Large Language Model Evaluatorscited by 146
  5. 2024The BiGGen Bench: A Principled Benchmark for Fine-grained Evaluation of Language Models with Language Modelscited by 89
  6. 2024Systematic Evaluation of LLM-as-a-Judge in LLM Alignment Tasks: Explainable Metrics and Diverse Prompt Templatescited by 71
  7. 2024Can You Trust LLM Judgments? Reliability of LLM-as-a-Judgecited by 52
  8. 2024HD-Eval: Aligning Large Language Model Evaluators Through Hierarchical Criteria Decompositioncited by 33
  9. 2024Finding Blind Spots in Evaluator LLMs with Interpretable Checklistscited by 32
  10. 2023Goal-Oriented Prompt Attack and Safety Evaluation for LLMscited by 23

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 4; 2026: 13.

17 papers.

Sort:

By area:

  1. 2025Evaluating LLM Agent Collusion in Double Auctionscited by 11
  2. 2026When the Judge Changes, So Does the Measurement: Auditing LLM-as-Judge Reliabilitycited by 6
  3. 2025The Artificial Intelligence Cognitive Examination: A Survey on the Evolution of Multimodal Evaluation from Recognition to Reasoningcited by 4
  4. 2025Uncovering Vulnerabilities of LLM-Assisted Cyber Threat Intelligencecited by 3
  5. 2026Building Customer Support AI Agents at 100M-User Scale: An Evaluation-Driven Frameworkcited by 3
  6. 2026ComplexConstraints and Beyond: Expert Rubrics for RLVRcited by 3
  7. 2026Omissive Bias in Religious Representation: Benchmarking LLM Answers to Everyday Ethical Decision-makingcited by 2
  8. 2026AI for Games in the Foundation Model Eracited by 2
  9. 2025CancerGUIDE: Cancer Guideline Understanding via Internal Disagreement Estimationcited by 1
  10. 2026Early-Token Confidence Predicts Reasoning Quality in Multi-Agent LLM Debatecited by 1
  11. 2026LLM-as-a-Judge Scores Are Unreliable Optimization Signals in Closed-Loop Table Recognitioncited by 1
  12. 2026Benchmarking Bengali Dialectal Bias: A Multi-Stage Framework Integrating RAG-Based Translation and Human-Augmented RLAIFcited by 0
5 more
  1. 2026Are LLMs Bad at Moral Reasoning?cited by 0
  2. 2026ToolLoop: Closed-Loop Tool-Use Data Synthesis via Decomposed Generation and Dynamic Self-Feedbackcited by 0
  3. 2026Beyond Accuracy and Surface Fluency: Risk-Sensitive Evaluation of LLMs for Legal Clause Generationcited by 0
  4. 2026From Dead Code and Static Requirements to Working Engines: Software Revival with Coding Agentscited by 0
  5. 2026Risk-Controlled Selective LLM Answering by Pricing Label-Free Checkscited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2024Aligning Human and LLM Judgments: Insights from EvalAssist on Task-Specific Evaluations and AI-assisted Assessment Strategy Preferences
  2. 2025LLM-as-a-Judge for Software Engineering: Literature Review, Vision, and the Road Ahead ↗
  3. 2024Enhancing LLM Evaluations: The Garbling Trick
  4. 2025From Code to Courtroom: LLMs as the New Software Judges
  5. 2025Analysis of instruction-based LLMs' capabilities to score and judge text-input problems in an academic setting
  6. 2024Evaluating the Evaluator: Measuring LLMs' Adherence to Task Evaluation Instructions
  7. 2025On Evaluating LLM Alignment by Evaluating LLMs as Judges
  8. 2025Training an LLM-as-a-Judge Model: Pipeline, Insights, and Practical Lessons