mlsanityData to 7 Oct 2026

Are LLM-Judges Robust to Expressions of Uncertainty? Investigating the effect of Epistemic Markers on LLM-based Evaluation

arXiv 2410.20774 ↗First posted 28 Oct 2024cs.CLcited by 29

Not on the map. Why?

Abstract (arXiv)

In line with the principle of honesty, there has been a growing effort to train large language models (LLMs) to generate outputs containing epistemic markers. However, evaluation in the presence of epistemic markers has been largely overlooked, raising a critical question: Could the use of epistemic markers in LLM-generated outputs lead to unintended negative consequences? To address this, we present EMBER, a benchmark designed to assess the robustness of LLM-judges to epistemic markers in both single and pairwise evaluation settings. Our findings, based on evaluations using EMBER, reveal that all tested LLM-judges, including GPT-4o, show a notable lack of robustness in the presence of epistemic markers. Specifically, we observe a negative bias toward epistemic markers, with a stronger bias against markers expressing uncertainty. This suggests that LLM-judges are influenced by the presence of these markers and do not focus solely on the correctness of the content.

Key facts. Cited by 29 papers in the map (data through 7 Oct 2026). Builds on 34 of its 61 references that are in the map. First posted on arXiv 28 Oct 2024.

Builds on

34 of 61 references.

References · 34

  1. 2023GPT-4 Technical Reportcited by 18,132
  2. 2024The Llama 3 Herd of Modelscited by 14,192
  3. 2023Judging LLM-as-a-Judge with MT-Bench and Chatbot Arenacited by 8,393
  4. 2017TriviaQA: A Large Scale Distantly Supervised Challenge Dataset for Reading Comprehensioncited by 2,895
  5. 2023G-Eval: NLG Evaluation using GPT-4 with Better Human Alignmentcited by 2,205
  6. 2022Language Models (Mostly) Know What They Knowcited by 1,783
  7. 2017Explanation in Artificial Intelligence: Insights from the Social Sciencescited by 1,201
  8. 2023Large Language Models are not Fair Evaluatorscited by 1,050
  9. 2021A General Language Assistant as a Laboratory for Alignmentcited by 1,023
  10. 2023Can LLMs Express Their Uncertainty? An Empirical Evaluation of Confidence Elicitation in LLMscited by 946
  11. 2023Just Ask for Calibration: Strategies for Eliciting Calibrated Confidence Scores from Language Models Fine-Tuned with Human Feedbackcited by 788
  12. 2023AlpacaFarm: A Simulation Framework for Methods that Learn from Human Feedbackcited by 738
22 more
  1. 2022Teaching Models to Express Their Uncertainty in Wordscited by 731
  2. 2023Can Large Language Models Be an Alternative to Human Evaluations?cited by 729
  3. 2023LLM-Blender: Ensembling Large Language Models with Pairwise Ranking and Generative Fusioncited by 550
  4. 2023OPERA: Alleviating Hallucination in Multi-Modal Large Language Models via Over-Trust Penalty and Retrospection-Allocationcited by 468
  5. 2023Trustworthy LLMs: a Survey and Guideline for Evaluating Large Language Models' Alignmentcited by 394
  6. 2024MLLM-as-a-Judge: Assessing Multimodal LLM-as-a-Judge with Vision-Language Benchmarkcited by 362
  7. 2023JudgeLM: Fine-tuned Large Language Models are Scalable Judgescited by 351
  8. 2023Evaluating Large Language Models at Evaluating Instruction Followingcited by 290
  9. 2024Judging the Judges: Evaluating Alignment and Vulnerabilities in LLMs-as-Judgescited by 197
  10. 2024Is LLM-as-a-Judge Robust? Investigating Universal Adversarial Attacks on Zero-shot LLM Assessmentcited by 117
  11. 2021Truthful AI: Developing and governing AI that does not liecited by 109
  12. 2024Relying on the Unreliable: The Impact of Language Models' Reluctance to Express Uncertaintycited by 106
  13. 2023Breaking Common Sense: WHOOPS! A Vision-and-Language Benchmark of Synthetic and Compositional Imagescited by 79
  14. 2023Automated Annotation with Generative AI Requires Validationcited by 75
  15. 2023Cognitive Dissonance: Why Do Language Model Outputs Disagree with Internal Representations of Truthfulness?cited by 62
  16. 2024MARS: Meaning-Aware Response Scoring for Uncertainty Estimation in Generative LLMscited by 45
  17. 2023ROME: Evaluating Pre-trained Vision-Language Models on Reasoning beyond Visual Common Sensecited by 27
  18. 2023A Diachronic Perspective on User Trust in AI under Uncertaintycited by 23
  19. 2023Missing Information, Unresponsive Authors, Experimental Flaws: The Impossibility of Assessing the Reproducibility of Previous Human Evaluations in NLPcited by 21
  20. 2024Mitigating Hallucinations in Large Vision-Language Models via Summary-Guided Decodingcited by 21
  21. 2024AdvisorQA: Towards Helpful and Harmless Advice-seeking Question Answering with Collective Intelligencecited by 12
  22. 2024Return of EM: Entity-driven Answer Set Expansion for QA Evaluationcited by 2

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2024: 1; 2025: 13; 2026: 15.

29 papers.

Sort:

By area:

  1. 2024JuStRank: Benchmarking LLM Judges for System Rankingcited by 27
  2. 2025Conformal Linguistic Calibration: Trading-off between Factuality and Specificitycited by 20
  3. 2025Don't Judge Code by Its Cover: Exploring Biases in LLM Judges for Code Evaluationcited by 18
  4. 2025Revisiting Epistemic Markers in Confidence Estimation: Can Markers Accurately Reflect Large Language Models' Uncertainty?cited by 16
  5. 2025Anthropomimetic Uncertainty: What Verbalized Uncertainty in Language Models is Missingcited by 14
  6. 2025Robustness in Large Language Models: A Survey of Mitigation Strategies and Evaluation Metricscited by 12
  7. 2025Fooling the LVLM Judges: Visual Biases in LVLM-Based Evaluationcited by 10
  8. 2025Can You Trick the Grader? Adversarial Persuasion of LLM Judgescited by 9
  9. 2026When Wording Steers the Evaluation: Framing Bias in LLM judgescited by 8
  10. 2025Rethinking Prospect Theory for LLMs: Revealing the Instability of Decision-Making under Epistemic Uncertaintycited by 3
  11. 2025EVALUESTEER: Measuring Reward Model Steerability Towards Values and Preferencescited by 3
  12. 2026Judging Against the Reference: Uncovering Knowledge-Driven Failures in LLM-Judges on QA Evaluationcited by 3
17 more
  1. 2026When Planning Fails Despite Correct Execution: On Epistemic Calibration for LLM-Based Multi-Agent Systemscited by 3
  2. 2026Can LLMs Use Linguistic Uncertainty Markers to Reliably Reflect Intrinsic Confidence?cited by 2
  3. 2025Single LLM Debate, MoLaCE: Mixture of Latent Concept Experts Against Confirmation Biascited by 1
  4. 2026Auditing Stealth Sycophancy in Mental-Health Dialogue: Structured Clinical-State Diagnostics and Clean Matched Benchmarkscited by 1
  5. 2026Agreement Overstates Evidence: Error Dependence in LLM Judge Consensuscited by 1
  6. 2025Advancing Uto-Aztecan Language Technologies: A Case Study on the Endangered Comanche Languagecited by 0
  7. 2025FinReflectKG -- EvalBench: Benchmarking Financial KG with Multi-Dimensional Evaluationcited by 0
  8. 2025Mapping Clinical Doubt: Locating Linguistic Uncertainty in LLMscited by 0
  9. 2026TARAZ: Persian Short-Answer Question Benchmark for Cultural Evaluation of Language Modelscited by 0
  10. 2026Saying More Than They Know: A Framework for Quantifying Epistemic-Rhetorical Miscalibration in Large Language Modelscited by 0
  11. 2026Faithful or Fabricated? A Causal Framework for Rationalization Bias in LLM Judgescited by 0
  12. 2026Framing Matters: Addressing Framing Sensitivity in Decision-Making through Behaviorally-Grounded Value Alignmentcited by 0
  13. 2026From 'May' to 'Is': Certainty Distortion in Language Model Rewritingcited by 0
  14. 2026Towards Spec Learning: Inference-Time Alignment from Preference Pairscited by 0
  15. 2026One More Turn, Less Regret: A Regret-Based Multi-Turn Benchmark for LLMs' Clarification Policiescited by 0
  16. 2026Silence Is Endorsement: Verification-Status Laundering in LLM Agent Pipelinescited by 0
  17. 2026A Missing Piece for Trustworthy AI Reviewers: From Benchmarking Rhetorical Robustness to SciCore Reviewcited by 0