mlsanityData to 7 Oct 2026

Can One-sided Arguments Lead to Response Change in Large Language Models?

arXiv 2602.06260 ↗First posted 5 Feb 2026cs.CL cs.AIcited by 0

Not on the map. Why?

Abstract (arXiv)

Polemic questions need more than one viewpoint to express a balanced answer. Large Language Models (LLMs) can provide a balanced answer, but also take a single aligned viewpoint or refuse to answer. In this paper, we study if such initial responses can be steered to a specific viewpoint in a simple and intuitive way: by only providing one-sided arguments supporting the viewpoint. Our systematic study has three dimensions: (i) which stance is induced in the LLM response, (ii) how the polemic question is formulated, (iii) how the arguments are shown. We construct a small dataset and remarkably find that opinion steering occurs across (i)-(iii) for diverse models, number of arguments, and topics. Switching to other arguments consistently decreases opinion steering.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Builds on 23 of its 37 references that are in the map. First posted on arXiv 5 Feb 2026.

Builds on

23 of 37 references.

References · 23

  1. 2023Mistral 7Bcited by 3,065
  2. 2023Towards Understanding Sycophancy in Language Modelscited by 988
  3. 2024Large Language Model based Multi-Agents: A Survey of Progress and Challengescited by 752
  4. 2022Using Large Language Models to Simulate Multiple Humans and Replicate Human Subject Studiescited by 506
  5. 2023XSTest: A Test Suite for Identifying Exaggerated Safety Behaviours in Large Language Modelscited by 477
  6. 2024Political Compass or Spinning Arrow? Towards More Meaningful Evaluations for Values and Opinions in Large Language Modelscited by 162
  7. 2023In-Context Impersonation Reveals Large Language Models' Strengths and Biasescited by 154
  8. 2024What Evidence Do Language Models Find Convincing?cited by 74
  9. 2018Community Interaction and Conflict on the Webcited by 72
  10. 2023Exploring the Potential of Large Language Models in Computational Argumentationcited by 42
  11. 2024Navigating the OverKill in Large Language Modelscited by 40
  12. 2024Can LLMs Speak For Diverse People? Tuning LLMs via Debate to Generate Controllable Controversial Statementscited by 33
11 more
  1. 2019Are You Convinced? Choosing the More Convincing Evidence with a Siamese Networkcited by 29
  2. 2024Characterizing LLM Abstention Behavior in Science QA with Context Perturbationscited by 26
  3. 2024The Potential and Challenges of Evaluating Attitudes, Opinions, and Values in Large Language Modelscited by 19
  4. 2024Which Side Are You On? A Multi-task Dataset for End-to-End Argument Summarisation and Evaluationcited by 13
  5. 2023DELPHI: Data for Evaluating LLMs' Performance in Handling Controversial Issuescited by 11
  6. 2025Large Language Models Are Better Logical Fallacy Reasoners with Counterargument, Explanation, and Goal-Aware Prompt Formulationcited by 10
  7. 2024A Logical Fallacy-Informed Framework for Argument Generationcited by 7
  8. 2024AutoPersuade: A Framework for Evaluating and Explaining Persuasive Argumentscited by 7
  9. 2024Large Language Models can Achieve Social Balancecited by 5
  10. 2023How to Handle Different Types of Out-of-Distribution Scenarios in Computational Argumentation? A Comprehensive and Fine-Grained Field Studycited by 4
  11. 2024Flee the Flaw: Annotating the Underlying Logic of Fallacious Arguments Through Templates and Slot-fillingcited by 2

Built on by

Not cited yet.