# Are LLM-based Evaluators Confusing NLG Quality Criteria?

arXiv: https://arxiv.org/abs/2402.12055 · HTML: https://mlsanity.com/paper/2402.12055/

## Key facts

- Cited by 36 papers in the map (data through 14 Oct 2026).
- Builds on 31 of its 49 references that are in the map.
- First posted on arXiv 19 Feb 2024.

## Builds on (most cited first, 30 of 31)

- [Llama 2: Open Foundation and Fine-Tuned Chat Models](https://mlsanity.com/paper/2307.09288/) (2023)
- [Judging LLM-as-a-Judge with MT-Bench and Chatbot Arena](https://mlsanity.com/paper/2306.05685/) (2023)
- [BERTScore: Evaluating Text Generation with BERT](https://mlsanity.com/paper/1904.09675/) (2019)
- [G-Eval: NLG Evaluation using GPT-4 with Better Human Alignment](https://mlsanity.com/paper/2303.16634/) (2023)
- [BLEURT: Learning Robust Metrics for Text Generation](https://mlsanity.com/paper/2004.04696/) (2020)
- [Large Language Models are not Fair Evaluators](https://mlsanity.com/paper/2305.17926/) (2023)
- [Teaching Machines to Read and Comprehend](https://mlsanity.com/paper/1506.03340/) (2015)
- [COMET: A Neural Framework for MT Evaluation](https://mlsanity.com/paper/2009.09025/) (2020)
- [Beyond Accuracy: Behavioral Testing of NLP models with CheckList](https://mlsanity.com/paper/2005.04118/) (2020)
- [Can Large Language Models Be an Alternative to Human Evaluations?](https://mlsanity.com/paper/2305.01937/) (2023)
- [SummEval: Re-evaluating Summarization Evaluation](https://mlsanity.com/paper/2007.12626/) (2020)
- [Evaluating the Factual Consistency of Abstractive Text Summarization](https://mlsanity.com/paper/1910.12840/) (2019)
- [Prometheus: Inducing Fine-grained Evaluation Capability in Language Models](https://mlsanity.com/paper/2310.08491/) (2023)
- [SAMSum Corpus: A Human-annotated Dialogue Dataset for Abstractive Summarization](https://mlsanity.com/paper/1911.12237/) (2019)
- [Is ChatGPT a Good NLG Evaluator? A Preliminary Study](https://mlsanity.com/paper/2303.04048/) (2023)
- [Large Language Models Are State-of-the-Art Evaluators of Translation Quality](https://mlsanity.com/paper/2302.14520/) (2023)
- [Neural Text Summarization: A Critical Evaluation](https://mlsanity.com/paper/1908.08960/) (2019)
- [USR: An Unsupervised and Reference Free Evaluation Metric for Dialog Generation](https://mlsanity.com/paper/2005.00456/) (2020)
- [Human-like Summarization Evaluation with ChatGPT](https://mlsanity.com/paper/2304.02554/) (2023)
- [Are Large Language Model-based Evaluators the Solution to Scaling Up Multilingual Evaluation?](https://mlsanity.com/paper/2309.07462/) (2023)
- [EvalLM: Interactive Evaluation of Large Language Model Prompts on User-Defined Criteria](https://mlsanity.com/paper/2309.13633/) (2023)
- [The Next Chapter: A Study of Large Language Models in Storytelling](https://mlsanity.com/paper/2301.09790/) (2023)
- [Evaluation Metrics in the Era of GPT-4: Reliably Evaluating Large Language Models on Sequence to Sequence Tasks](https://mlsanity.com/paper/2310.13800/) (2023)
- [Perturbation CheckLists for Evaluating NLG Evaluation Metrics](https://mlsanity.com/paper/2109.05771/) (2021)
- [On the Blind Spots of Model-Based Evaluation Metrics for Text Generation](https://mlsanity.com/paper/2212.10020/) (2022)
- [A Comprehensive Analysis of the Effectiveness of Large Language Models as Automatic Dialogue Evaluators](https://mlsanity.com/paper/2312.15407/) (2023)
- [DEMETR: Diagnosing Evaluation Metrics for Translation](https://mlsanity.com/paper/2210.13746/) (2022)
- [X-Eval: Generalizable Multi-aspect Text Evaluation via Augmented Instruction Tuning with Auxiliary Evaluation Aspects](https://mlsanity.com/paper/2311.08788/) (2023)
- [Simple LLM Prompting is State-of-the-Art for Robust and Multilingual Dialogue Evaluation](https://mlsanity.com/paper/2308.16797/) (2023)
- [CoAScore: Chain-of-Aspects Prompting for NLG Evaluation](https://mlsanity.com/paper/2312.10355/) (2023)

## Built on by (most cited first, 30 of 36)

- [From Generation to Judgment: Opportunities and Challenges of LLM-as-a-judge](https://mlsanity.com/paper/2411.16594/) (2024)
- [Judging the Judges: Evaluating Alignment and Vulnerabilities in LLMs-as-Judges](https://mlsanity.com/paper/2406.12624/) (2024)
- [Lessons from the Trenches on Reproducible Evaluation of Language Models](https://mlsanity.com/paper/2405.14782/) (2024)
- [Neither Valid nor Reliable? Investigating the Use of LLMs as Judges](https://mlsanity.com/paper/2508.18076/) (2025)
- [Toward Generalizable Evaluation in the LLM Era: A Survey Beyond Benchmarks](https://mlsanity.com/paper/2504.18838/) (2025)
- [Improving Context-Aware Preference Modeling for Language Models](https://mlsanity.com/paper/2407.14916/) (2024)
- [Reasoning Through Execution: Unifying Process and Outcome Rewards for Code Generation](https://mlsanity.com/paper/2412.15118/) (2024)
- [RevisEval: Improving LLM-as-a-Judge via Response-Adapted References](https://mlsanity.com/paper/2410.05193/) (2024)
- [Themis: A Reference-free NLG Evaluation Language Model with Flexibility and Interpretability](https://mlsanity.com/paper/2406.18365/) (2024)
- [Beyond the Surface: Enhancing LLM-as-a-Judge Alignment with Human via Internal Representations](https://mlsanity.com/paper/2508.03550/) (2025)
- [DHP Benchmark: Are LLMs Good NLG Evaluators?](https://mlsanity.com/paper/2408.13704/) (2024)
- [Decision Information Meets Large Language Models: The Future of Explainable Operations Research](https://mlsanity.com/paper/2502.09994/) (2025)
- [Rubrik's Cube: Testing a New Rubric for Evaluating Explanations on the CUBE dataset](https://mlsanity.com/paper/2503.23899/) (2025)
- [Multi-Agent LLM Judge: automatic personalized LLM judge design for evaluating natural language generation applications](https://mlsanity.com/paper/2504.02867/) (2025)
- [CLEV: LLM-Based Evaluation Through Lightweight Efficient Voting for Free-Form Question-Answering](https://mlsanity.com/paper/2503.08542/) (2025)
- [HPSS: Heuristic Prompting Strategy Search for LLM Evaluators](https://mlsanity.com/paper/2502.13031/) (2025)
- [Exploring the Multilingual NLG Evaluation Abilities of LLM-Based Evaluators](https://mlsanity.com/paper/2503.04360/) (2025)
- [Auto-Prompt Ensemble for LLM Judge](https://mlsanity.com/paper/2510.06538/) (2025)
- [OpeNLGauge: An Explainable Metric for NLG Evaluation with Open-Weights LLMs](https://mlsanity.com/paper/2503.11858/) (2025)
- [Revisiting NLI: Towards Cost-Effective and Human-Aligned Metrics for Evaluating LLMs in Question Answering](https://mlsanity.com/paper/2511.07659/) (2025)
- [Focal Reward: Balanced Reinforcement Learning under Rubric-Based Rewards](https://mlsanity.com/paper/2605.26579/) (2026)
- [Meta-Judging with Large Language Models: Concepts, Methods, and Challenges](https://mlsanity.com/paper/2601.17312/) (2026)
- [Useless but Safe? Benchmarking Utility Recovery with User Intent Clarification in Multi-Turn Conversations](https://mlsanity.com/paper/2604.27093/) (2026)
- [EvolvR: Self-Evolving Pairwise Reasoning for Story Evaluation to Enhance Generation](https://mlsanity.com/paper/2508.06046/) (2025)
- [Calibrating MLLM-as-a-judge via Multimodal Bayesian Prompt Ensembles](https://mlsanity.com/paper/2509.08777/) (2025)
- [Reward Models are Metrics in a Trench Coat](https://mlsanity.com/paper/2510.03231/) (2025)
- [Reflections and New Directions for Human-Centered Large Language Models](https://mlsanity.com/paper/2605.06901/) (2026)
- [A Dual-Perspective NLG Meta-Evaluation Framework with Automatic Benchmark and Better Interpretability](https://mlsanity.com/paper/2502.12052/) (2025)
- [What Are We Measuring in NLG? A Meta-Analysis of Evaluation Trends 2020-2025](https://mlsanity.com/paper/2601.07648/) (2026)
- [Graph-Structured Rubrics: Compiling Rubrics into Typed Evaluation Graphs for LLM Judges](https://mlsanity.com/paper/2608.12097/) (2026)

## Abstract (arXiv)

Some prior work has shown that LLMs perform well in NLG evaluation for different tasks. However, we discover that LLMs seem to confuse different evaluation criteria, which reduces their reliability. For further verification, we first consider avoiding issues of inconsistent conceptualization and vague expression in existing NLG quality criteria themselves. So we summarize a clear hierarchical classification system for 11 common aspects with corresponding different criteria from previous studies involved. Inspired by behavioral testing, we elaborately design 18 types of aspect-targeted perturbation attacks for fine-grained analysis of the evaluation behaviors of different LLMs. We also conduct human annotations beyond the guidance of the classification system to validate the impact of the perturbations. Our experimental results reveal confusion issues inherent in LLMs, as well as other noteworthy phenomena, and necessitate further research and improvements for LLM-based evaluation.