Probing Association Biases in LLM Moderation Over-Sensitivity

• URL: http://arxiv.org/abs/2505.23914v1
• Date: Thu, 29 May 2025 18:07:48 GMT
• Title: Probing Association Biases in LLM Moderation Over-Sensitivity
• Authors: Yuxin Wang, Botao Yu, Ivory Yang, Saeed Hassanpour, Soroush Vosoughi,
• Abstract summary: Large Language Models are widely used for content moderation but often misclassify benign comments as toxic.<n>We introduce Topic Association Analysis, a semantic-level approach to quantify how LLMs associate certain topics with toxicity.<n>More advanced models (e.g., GPT-4 Turbo) demonstrate stronger topic stereotype despite lower overall false positive rates.
• Score: 42.191744175730726
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Large Language Models are widely used for content moderation but often misclassify benign comments as toxic, leading to over-sensitivity. While previous research attributes this issue primarily to the presence of offensive terms, we reveal a potential cause beyond token level: LLMs exhibit systematic topic biases in their implicit associations. Inspired by cognitive psychology's implicit association tests, we introduce Topic Association Analysis, a semantic-level approach to quantify how LLMs associate certain topics with toxicity. By prompting LLMs to generate free-form scenario imagination for misclassified benign comments and analyzing their topic amplification levels, we find that more advanced models (e.g., GPT-4 Turbo) demonstrate stronger topic stereotype despite lower overall false positive rates. These biases suggest that LLMs do not merely react to explicit, offensive language but rely on learned topic associations, shaping their moderation decisions. Our findings highlight the need for refinement beyond keyword-based filtering, providing insights into the underlying mechanisms driving LLM over-sensitivity.

Related papers

• What's Taboo for You? - An Empirical Evaluation of LLMs Behavior Toward Sensitive Content [1.6492989697868894]
  This study empirically analyzes the implicit moderation behavior of GPT-4o-mini when paraphrasing sensitive content.<n>Our experiments indicate that GPT-4o-mini systematically moderates content toward less sensitive classes, with substantial reductions in derogatory and taboo language.
  arXiv  Detail & Related papers  (2025-07-31T08:02:04Z)
• Beyond Keywords: Evaluating Large Language Model Classification of Nuanced Ableism [2.0435202333125977]
  Large language models (LLMs) are increasingly used in decision-making tasks like r'esum'e screening and content moderation.<n>We evaluate the ability of four LLMs to identify nuanced ableism directed at autistic individuals.<n>Our results reveal that LLMs can identify autism-related language but often miss harmful or offensive connotations.
  arXiv  Detail & Related papers  (2025-05-26T20:01:44Z)
• Information Anxiety in Large Language Models [21.574677910096735]
  Large Language Models (LLMs) have demonstrated strong performance as knowledge repositories. We take the investigation further by conducting a comprehensive analysis of the internal reasoning and retrieval mechanisms of LLMs. Our work focuses on three critical dimensions - the impact of entity popularity, the models' sensitivity to lexical variations in query formulation, and the progression of hidden state representations.
  arXiv  Detail & Related papers  (2024-11-16T14:28:33Z)
• Hate Personified: Investigating the role of LLMs in content moderation [64.26243779985393]
  For subjective tasks such as hate detection, where people perceive hate differently, the Large Language Model's (LLM) ability to represent diverse groups is unclear. By including additional context in prompts, we analyze LLM's sensitivity to geographical priming, persona attributes, and numerical information to assess how well the needs of various groups are reflected.
  arXiv  Detail & Related papers  (2024-10-03T16:43:17Z)
• Analyzing LLM Behavior in Dialogue Summarization: Unveiling Circumstantial Hallucination Trends [38.86240794422485]
  We evaluate the faithfulness of large language models for dialogue summarization. Our evaluation reveals subtleties as to what constitutes a hallucination. We introduce two prompt-based approaches for fine-grained error detection that outperform existing metrics.
  arXiv  Detail & Related papers  (2024-06-05T17:49:47Z)
• Don't Go To Extremes: Revealing the Excessive Sensitivity and Calibration Limitations of LLMs in Implicit Hate Speech Detection [29.138463029748547]
  This paper explores the capability of Large Language Models to detect implicit hate speech and express confidence in their responses. Our findings highlight that LLMs exhibit two extremes: (1) LLMs display excessive sensitivity towards groups or topics that may cause fairness issues, resulting in misclassifying benign statements as hate speech.
  arXiv  Detail & Related papers  (2024-02-18T00:04:40Z)
• A Theory of LLM Sampling: Part Descriptive and Part Prescriptive [53.08398658452411]
  Large Language Models (LLMs) are increasingly utilized in autonomous decision-making.<n>We show that this sampling behavior resembles that of human decision-making.<n>We show that this deviation of a sample from the statistical norm towards a prescriptive component consistently appears in concepts across diverse real-world domains.
  arXiv  Detail & Related papers  (2024-02-16T18:28:43Z)
• Exploring the Jungle of Bias: Political Bias Attribution in Language Models via Dependency Analysis [86.49858739347412]
  Large Language Models (LLMs) have sparked intense debate regarding the prevalence of bias in these models and its mitigation. We propose a prompt-based method for the extraction of confounding and mediating attributes which contribute to the decision process. We find that the observed disparate treatment can at least in part be attributed to confounding and mitigating attributes and model misalignment.
  arXiv  Detail & Related papers  (2023-11-15T00:02:25Z)
• Are Large Language Models Really Robust to Word-Level Perturbations? [68.60618778027694]
  We propose a novel rational evaluation approach that leverages pre-trained reward models as diagnostic tools. Longer conversations manifest the comprehensive grasp of language models in terms of their proficiency in understanding questions. Our results demonstrate that LLMs frequently exhibit vulnerability to word-level perturbations that are commonplace in daily language usage.
  arXiv  Detail & Related papers  (2023-09-20T09:23:46Z)
• Siren's Song in the AI Ocean: A Survey on Hallucination in Large Language Models [116.01843550398183]
  Large language models (LLMs) have demonstrated remarkable capabilities across a range of downstream tasks. LLMs occasionally generate content that diverges from the user input, contradicts previously generated context, or misaligns with established world knowledge.
  arXiv  Detail & Related papers  (2023-09-03T16:56:48Z)
• Language models are not naysayers: An analysis of language models on negation benchmarks [58.32362243122714]
  We evaluate the ability of current-generation auto-regressive language models to handle negation. We show that LLMs have several limitations including insensitivity to the presence of negation, an inability to capture the lexical semantics of negation, and a failure to reason under negation.
  arXiv  Detail & Related papers  (2023-06-14T01:16:37Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.