Lune

EMNLP Findings2025

Curse of Knowledge: Your Guidance and Provided Knowledge are biasing LLM Judges in Complex Evaluation

Weiyuan Li, Xintao Wang, Siyu Yuan, Rui Xu, Jiangjie Chen, Qingqing Dong, Yanghua Xiao, Deqing Yang

Abstract

As large language models (LLMs) grow more capable, they face increasingly diverse and complex tasks, making reliable evaluation challenging. The paradigm of LLMs as judges has emerged as a scalable solution, yet prior work primarily focuses on simple settings. Their reliability in complex tasks-where multi-faceted rubrics, unstructured reference answers, and nuanced criteria are critical-remains understudied. In this paper, we constructed COM-PLEXEVAL, a challenge benchmark designed to systematically expose and quantify Auxiliary Information Induced Biases. We systematically investigated and validated 6 previously unexplored biases across 12 basic and 3 advanced scenarios. Key findings reveal: (1) all evaluated models exhibit significant susceptibility to these biases, with bias magnitude scaling with task complexity; (2) notably, Large Reasoning Models (LRMs) show paradoxical vulnerability. Our in-depth analysis offers crucial insights for improving the accuracy and verifiability of evaluation signals, paving the way for more general and robust evaluation models.

Ask Lune

Know this paper in minutes, not an afternoon.

Lune can read every page of it and the top-tier papers that cite it, and gives you its claims, its evidence and its limits, each one cited to the paper it comes from. No skimming, no AI slop.

Questions to start from

Lune can draw on

the full paper · related work

Get the answer

Free to start. No credit card required.

Lineage

  • Self-Rewarding Language Models cites Prometheus: Inducing Fine-Grained Evaluation Capability in Language Models
  • ChatEval: Towards Better LLM-based Evaluators through Multi-Agent Debate cites Language Models are Few-Shot Learners
Builds onCited byLeft to right by year; the more cited, the higher.

Builds on28

  1. Language Models are Few-Shot LearnersTom B. Brown, Benjamin Mann, Nick Ryder, Melanie Subbiah et al.NeurIPS 2020 · 64,255 citations
  2. ChatEval: Towards Better LLM-based Evaluators through Multi-Agent DebateChi-Min Chan, Weize Chen, Yusheng Su, Jianxuan Yu et al.ICLR 2024 · 871 citations
  3. Self-Rewarding Language ModelsWeizhe Yuan, Richard Yuanzhe Pang, Kyunghyun Cho, Xian Li et al.ICML 2024 · 569 citations
  4. Prometheus: Inducing Fine-Grained Evaluation Capability in Language ModelsSeungone Kim, Jamin Shin, Yejin Choi, Joel Jang et al.ICLR 2024 · 468 citations
  5. ProcessBench: Identifying Process Errors in Mathematical ReasoningChujie Zheng, Zhenru Zhang, Beichen Zhang, Runji Lin et al.ACL 2025 · 209 citations

Related papers

  1. MM-JudgeBias: A Benchmark for Evaluating Compositional Biases in MLLM-as-a-JudgeSua Lee, Sanghee Park, Jinbae ImACL 2026 · 1 citation
  2. RubricBench: Aligning Model-Generated Rubrics with Human StandardsJunyi Zhou, Qiyuan Zhang, Yufei Wang, Fuyuan Lyu et al.ACL 2026 · 7 citations
  3. Justice or Prejudice? Quantifying Biases in LLM-as-a-JudgeJiayi Ye, Yanbo Wang, Yue Huang, Dongping Chen et al.ICLR 2025
  4. LexInstructEval: Lexical Instruction Following Evaluation for Large Language ModelsHuimin Ren, Yan Liang, Baiqiao Su, Chaobo Sun et al.AAAI 2026
  5. CRUXEVAL-X: A Benchmark for Multilingual Code Reasoning, Understanding and ExecutionRuiyang Xu, Jialun Cao, Yaojie Lu, Ming Wen et al.ACL 2025 · 26 citations
Dusk over the sea between two cliffs drawn in fine vertical lines