mlsanityData to 14 Oct 2026

Narrowing the Horizon: Quantifying Topic Saliency Shifts in Generative Monoculture

arXiv 2609.35302 ↗First posted 28 Sep 2026cs.AIcited by 0

Not on the map. Why?

Abstract (arXiv)

As Large Language Models (LLMs) become central to how we access and share information, they play an increasingly powerful role in shaping global knowledge. However, as these models evolve, their outputs risk converging into a textitgenerative monoculture, where the diversity of perspectives they represent narrows over time. Studies at the model level often fail to pinpoint which specific topics or viewpoints are being marginalised or amplified in this process. In this paper, we introduce a method to measure shifts in topic saliency across model families, tracking what gains or loses prominence during post-training. Applying this approach to a case study of climate change discourse, we demonstrate how homogenisation affects the representation of diverse solutions across different models. We also test interventions to counter this trend, showing that specialised models can help preserve a broader range of perspectives. This underscores the importance of monitoring topic saliency to diagnose the risks of monoculture and to ensure AI systems reflect a pluralism of ideas. Data and Code are accessible hrefhttps://github.com/oriane/topicsaliencyshifthere.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 39 of its 89 references that are in the map. First posted on arXiv 28 Sep 2026.

Builds on

39 of 89 references.

References · 39

  1. 2019The Curious Case of Neural Text Degenerationcited by 2,916
  2. 2023Whose Opinions Do Language Models Reflect?cited by 773
  3. 2023Towards Measuring the Representation of Subjective Global Opinions in Language Modelscited by 437
  4. 2023Understanding the Effects of RLHF on LLM Generalisation and Diversitycited by 396
  5. 2023From Pretraining Data to Language Models to Downstream Tasks: Tracking the Trails of Political Biases Leading to Unfair NLP Modelscited by 260
  6. 2023Self-Consuming Generative Models Go MADcited by 250
  7. 2023Does Writing with Language Models Reduce Content Diversity?cited by 192
  8. 2023Generative artificial intelligence enhances creativity but reduces the diversity of novel contentcited by 167
  9. 2024Political Compass or Spinning Arrow? Towards More Meaningful Evaluations for Values and Opinions in Large Language Modelscited by 162
  10. 2024Large language models that replace human participants can harmfully misportray and flatten identity groupscited by 144
  11. 2025Artificial Hivemind: The Open-Ended Homogeneity of Language Models (and Beyond)cited by 141
  12. 2024Homogenization Effects of Large Language Models on Human Creative Ideationcited by 127
27 more
  1. 2024Mapping the Increasing Use of LLMs in Scientific Paperscited by 118
  2. 2024A Note on Shumailov et al. (2024): `AI Models Collapse When Trained on Recursively Generated Data'cited by 115
  3. 2024Modular Pluralism: Pluralistic Alignment via Multi-LLM Collaborationcited by 113
  4. 2023The Curious Decline of Linguistic Diversity: Training Language Models on Synthetic Textcited by 100
  5. 2024Bias and Unfairness in Information Retrieval Systems: New Challenges in the LLM Eracited by 98
  6. 2022Picking on the Same Person: Does Algorithmic Monoculture lead to Outcome Homogenization?cited by 91
  7. 2023Personalisation within bounds: A risk taxonomy and policy framework for the alignment of large language models with personalised feedbackcited by 87
  8. 2024AI Suggestions Homogenize Writing Toward Western Styles and Diminish Cultural Nuancescited by 60
  9. 2024One fish, two fish, but not the whole sea: Alignment reduces language models' conceptual diversitycited by 58
  10. 2025Large Language Model Hacking: Quantifying the Hidden Risks of Using LLMs for Text Annotationcited by 50
  11. 2025The Homogenizing Effect of Large Language Models on Human Expression and Thoughtcited by 35
  12. 2024AI and the Problem of Knowledge Collapsecited by 34
  13. 2023Can Large Language Models Capture Dissenting Human Voices?cited by 33
  14. 2025Position: Model Collapse Does Not Mean What You Thinkcited by 33
  15. 2025IssueBench: Millions of Realistic Prompts for Measuring Issue Bias in LLM Writing Assistancecited by 26
  16. 2024Detecting Mode Collapse in Language Models via Narrationcited by 24
  17. 2024The Rise of AI-Generated Content in Wikipediacited by 22
  18. 2025VITAL: A New Dataset for Benchmarking Pluralistic Alignment in Healthcarecited by 17
  19. 2024Prompting Diverse Ideas: Increasing AI Idea Variancecited by 14
  20. 2023Decolonial AI Alignment: Openness, Vi'sedsa-Dharma, and Including Excluded Knowledgescited by 10
  21. 2024To Diverge or Not to Diverge: A Morphosyntactic Perspective on Machine Translation vs Human Translationcited by 10
  22. 2022An Analysis of the Effects of Decoding Algorithms on Fairness in Open-Ended Language Generationcited by 7
  23. 2025Decentralising LLM Alignment: A Case for Context, Pluralism, and Participationcited by 5
  24. 2024Position: Cracking the Code of Cascading Disparity Towards Marginalized Communitiescited by 4
  25. 2024Understanding The Effect Of Temperature On Alignment With Human Opinionscited by 4
  26. 2026Reactive Writers: How Co-Writing with AI Changes How We Engage with Ideascited by 4
  27. 2026Ambiguity Collapse by LLMs: A Taxonomy of Epistemic Riskscited by 3

Built on by

Not cited yet.