More Is Not More: What Matters for Diversity in LLM Opinions?

• URL: http://arxiv.org/abs/2607.20429v1
• Date: Sun, 10 May 2026 00:52:25 GMT
• Title: More Is Not More: What Matters for Diversity in LLM Opinions?
• Abstract summary: We design a factorial experiment that separates two primary intervention dimensions: input conditioning and interaction architecture.<n>We evaluate all conditions on 100 real-user open-ended questions across 7 models, measuring diversity with multiple complementary metrics.<n>Our work demonstrates that diversity is not a product of scaling along any single dimension, but is highly sensitive to the structural form and combination of interventions.
• Score: 0.0
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Large language models are increasingly used to simulate diverse human opinions in open-ended tasks such as synthetic surveys, focus group modeling, and public opinion prediction. However, LLM outputs exhibit systematic opinion homogenization. Practitioners have explored various interventions to increase diversity, but the landscape remains fragmented: different methods are evaluated in isolation with incomparable metrics, and in practice they are typically deployed and upgraded simultaneously, making it difficult to attribute gains to specific components. To advance a more scientific understanding of LLM output diversity, we design a factorial experiment that separates two primary intervention dimensions: input conditioning (operationalized through persona depth) and interaction architecture. We evaluate all conditions on 100 real-user open-ended questions across 7 models, measuring diversity with multiple complementary metrics. Our findings challenge several common assumptions. First, more persona detail does not monotonically increase diversity. The initial step of persona conditioning already captures the majority of the gain, while further elaboration with demographic detail does not consistently improve and can reduce diversity on some models. Second, rather than seeking a single best interaction architecture, we find that different architectures explore largely non-overlapping opinion regions. Combining multiple architectures yields broader coverage than optimizing any one. Third, commonly attempted low-cost alternatives such as raising sampling temperature and adding diversity instructions produce negligible effects compared to structured interventions. Overall, our work demonstrates that diversity is not a product of scaling along any single dimension, but is highly sensitive to the structural form and combination of interventions.

Related papers

• Diversity Collapse in Multi-Agent LLM Systems: Structural Coupling and Collective Failure in Open-Ended Idea Generation [38.13720312846963]
  We study diversity in MAS-based ideation across three bottom-up levels: model intelligence, agent cognition, and system dynamics.<n>We characterize these outcomes as collective failures emerging from structural coupling.<n>Our analysis shows that this collapse arises primarily from the interaction structure rather than inherent model insufficiency.
  arXiv  Detail & Related papers  (2026-04-20T09:27:49Z)
• pMoE: Prompting Diverse Experts Together Wins More in Visual Adaptation [68.3777121585281]
  We propose a novel Mixture-of-Experts prompt tuning method called pMoE.<n>The proposed pMoE significantly enhances the model's versatility and applicability to a broad spectrum of tasks.<n>We conduct extensive experiments across 47 adaptation tasks, including both classification and segmentation in general and medical domains.
  arXiv  Detail & Related papers  (2026-02-26T12:27:06Z)
• Artificial Hivemind: The Open-Ended Homogeneity of Language Models (and Beyond) [90.45301024940329]
  Language models (LMs) often struggle to generate diverse, human-like creative content.<n>We introduce Infinity-Chat, a large-scale dataset of 26K diverse, real-world, open-ended user queries.<n>We present a large-scale study of mode collapse in LMs, revealing a pronounced Artificial Hivemind effect.
  arXiv  Detail & Related papers  (2025-10-27T03:16:21Z)
• Mind the Gap: Conformative Decoding to Improve Output Diversity of Instruction-Tuned Large Language Models [0.0]
  This paper investigates the diversity gap'' for a writing prompt narrative generation task.<n>Results show significant decreases in diversity due to instruction-tuning.<n>We present a new decoding strategy, conformative decoding, which guides an instruct model using its more diverse base model to reintroduce output diversity.
  arXiv  Detail & Related papers  (2025-07-28T16:04:25Z)
• Diverse LLMs or Diverse Question Interpretations? That is the Ensembling Question [5.847084649531299]
  We compare two diversity approaches for answering binary questions using large language models (LLMs)<n>For both cases, we apply majority voting as the ensemble consensus to determine the final answer.<n>Experiments on boolq, strategyqa, and pubmedqa show that question interpretation diversity consistently leads to better ensemble accuracy.
  arXiv  Detail & Related papers  (2025-07-25T15:26:18Z)
• Evaluating the Diversity and Quality of LLM Generated Content [72.84945252821908]
  We introduce a framework for measuring effective semantic diversity--diversity among outputs that meet quality thresholds.<n>Although preference-tuned models exhibit reduced lexical and syntactic diversity, they produce greater effective semantic diversity than SFT or base models.<n>These findings have important implications for applications that require diverse yet high-quality outputs.
  arXiv  Detail & Related papers  (2025-04-16T23:02:23Z)
• Exploring and Controlling Diversity in LLM-Agent Conversation [13.69653913986299]
  We propose Adaptive Prompt Pruning (APP), a novel method that allows users to control diversity via a single parameter.<n>APP effectively modulates diversity through extensive experiments and is compatible with existing diversity control methods.
  arXiv  Detail & Related papers  (2024-12-30T17:25:58Z)
• Scaling Data Diversity for Fine-Tuning Language Models in Human Alignment [84.32768080422349]
  Alignment with human preference prevents large language models from generating misleading or toxic content. We propose a new formulation of prompt diversity, implying a linear correlation with the final performance of LLMs after fine-tuning.
  arXiv  Detail & Related papers  (2024-03-17T07:08:55Z)
• Identifiability Results for Multimodal Contrastive Learning [72.15237484019174]
  We show that it is possible to recover shared factors in a more general setup than the multi-view setting studied previously. Our work provides a theoretical basis for multimodal representation learning and explains in which settings multimodal contrastive learning can be effective in practice.
  arXiv  Detail & Related papers  (2023-03-16T09:14:26Z)
• Picking on the Same Person: Does Algorithmic Monoculture lead to Outcome Homogenization? [90.35044668396591]
  A recurring theme in machine learning is algorithmic monoculture: the same systems, or systems that share components, are deployed by multiple decision-makers. We propose the component-sharing hypothesis: if decision-makers share components like training data or specific models, then they will produce more homogeneous outcomes. We test this hypothesis on algorithmic fairness benchmarks, demonstrating that sharing training data reliably exacerbates homogenization. We conclude with philosophical analyses of and societal challenges for outcome homogenization, with an eye towards implications for deployed machine learning systems.
  arXiv  Detail & Related papers  (2022-11-25T09:33:11Z)
• Feature diversity in self-supervised learning [0.0]
  We investigate how these factors may affect overall generalization performance in the context of self-supervised learning with CNN models. We found that the last layer is the most diversified throughout the training. While the model's test error decreases with increasing epochs, its diversity drops.
  arXiv  Detail & Related papers  (2022-09-02T21:34:11Z)

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.