mlsanityData to 14 Oct 2026

I Think, Therefore I Am Under-Qualified? A Benchmark for Evaluating Linguistic Shibboleth Detection in LLM Hiring Evaluations

arXiv 2508.04939 ↗First posted 6 Aug 2025cs.CLcited by 1

On the map: LLM reasoning → Reasoning benchmarks and evaluation

Abstract (arXiv)

This paper introduces a comprehensive benchmark for evaluating how Large Language Models (LLMs) respond to linguistic shibboleths: subtle linguistic markers that can inadvertently reveal demographic attributes such as gender, social class, or regional background. Through carefully constructed interview simulations using 100 validated question-response pairs, we demonstrate how LLMs systematically penalize certain linguistic patterns, particularly hedging language, despite equivalent content quality. Our benchmark generates controlled linguistic variations that isolate specific phenomena while maintaining semantic equivalence, which enables the precise measurement of demographic bias in automated evaluation systems. We validate our approach along multiple linguistic dimensions, showing that hedged responses receive 25.6% lower ratings on average, and demonstrate the benchmark's effectiveness in identifying model-specific biases. This work establishes a foundational framework for detecting and measuring linguistic discrimination in AI systems, with broad applications to fairness in automated decision-making contexts.

Key facts. Cited by 1 paper in the map (data through 14 Oct 2026). Placed in Reasoning benchmarks and evaluation, one of 17 directions in LLM reasoning. Builds on 26 of its 226 references that are in the map. First posted on arXiv 6 Aug 2025.

Builds on

26 of 226 references.

References · 26

  1. 2020Language Models are Few-Shot Learnerscited by 32,419
  2. 2021On the Opportunities and Risks of Foundation Modelscited by 4,033
  3. 2016Equality of Opportunity in Supervised Learningcited by 2,007
  4. 2019A Survey on Bias and Fairness in Machine Learningcited by 1,917
  5. 2017Towards A Rigorous Science of Interpretable Machine Learningcited by 1,416
  6. 2016Man is to Computer Programmer as Woman is to Homemaker? Debiasing Word Embeddingscited by 1,312
  7. 2020RealToxicityPrompts: Evaluating Neural Toxic Degeneration in Language Modelscited by 1,255
  8. 2016Fair prediction with disparate impact: A study of bias in recidivism prediction instrumentscited by 672
  9. 2019Learning the Difference that Makes a Difference with Counterfactually-Augmented Datacited by 439
  10. 2019Nuanced Metrics for Measuring Unintended Bias with Real Data for Text Classificationcited by 404
  11. 2017Algorithmic decision making and the cost of fairnesscited by 403
  12. 2023Larger language models do in-context learning differentlycited by 394
14 more
  1. 2022Repairing the Cracked Foundation: A Survey of Obstacles in Evaluation Practices for Generated Textcited by 155
  2. 2017Fairness in Machine Learning: Lessons from Political Philosophycited by 149
  3. 2019On Measuring and Mitigating Biased Inferences of Word Embeddingscited by 112
  4. 2021Measure and Improve Robustness in NLP Models: A Surveycited by 107
  5. 2020Saving Face: Investigating the Ethical Concerns of Facial Recognition Auditingcited by 89
  6. 2020Multi-Dimensional Gender Bias Classificationcited by 80
  7. 2021Anticipating Safety Issues in E2E Conversational AI: Framework and Toolingcited by 78
  8. 2019Perturbation Sensitivity Analysis to Detect Unintended Model Biasescited by 74
  9. 2024Linguistic Bias in ChatGPT: Language Models Reinforce Dialect Discriminationcited by 56
  10. 2024Towards Implicit Bias Detection and Mitigation in Multi-Agent LLM Interactionscited by 51
  11. 2024Dialect prejudice predicts AI decisions about people's character, employability, and criminalitycited by 37
  12. 2014Keyword and Keyphrase Extraction Using Centrality Measures on Collocation Networkscited by 2
  13. 2024Training LLMs to Recognize Hedges in Spontaneous Narrativescited by 1
  14. 2025Algorithmic Hiring and Diversity: Reducing Human-Algorithm Similarity for Better Outcomescited by 1

Built on by

Cited by year

Citing papers per year: 2025: 1.

1 paper.

  1. 2025Who's Your Judge? On the Detectability of LLM-Generated Judgmentscited by 8

Similar, not cited

Close in wording, no citation either way

  1. 2025Robustly Improving LLM Fairness in Realistic Settings via Interpretability
  2. 2025Investigating Intersectional Bias in Large Language Models using Confidence Disparities in Coreference Resolution
  3. 2024Decoding Biases: Automated Methods and LLM Judges for Gender Bias Detection in Language Models
  4. 2026Linguistic Triggers of Gender and Racial Bias in Open-Weight LLMs Applied to Recruitment
  5. 2024Interpreting Bias in Large Language Models: A Feature-Based Approach
  6. 2026Biases in the Blind Spot: Detecting What LLMs Fail to Mention
  7. 2025Dissecting Bias in LLMs: A Mechanistic Interpretability Perspective
  8. 2024BiasAlert: A Plug-and-play Tool for Social Bias Detection in LLMs