paper-with-me

Papers

How Susceptible are Large Language Models to Ideological Manipulation?

2024-02-18 · Kai Chen, Zihao He, Jun Yan, Taiwei Shi, Kristina Lerman

Large Language Models (LLMs) possess the potential to exert substantial influence on public perceptions and interactions with information. This raises concerns about the societal impact that could arise if the ideologies within these models can be easily manipulated. In this work, we investigate how effectively LLMs can learn and generalize ideological biases from their instruction-tuning data. Our findings reveal a concerning vulnerability: exposure to only a small amount of ideologically driven samples significantly alters the ideology of LLMs. Notably, LLMs demonstrate a startling ability to absorb ideology from one topic and generalize it to even unrelated ones. The ease with which LLMs' ideologies can be skewed underscores the risks associated with intentionally poisoned training data by malicious actors or inadvertently introduced biases by data annotators. It also emphasizes the imperative for robust safeguards to mitigate the influence of ideological manipulations on LLMs.

📄 PDF Abstract BibTeX arXiv:2402.11725

Code (1)

kaichen23/llm_ideo_manipulate 공식 구현

Similar Papers 제목 키워드 기반

Probing the Subtle Ideological Manipulation of Large Language Models

2025-04-19 · Demetris Paschalides, George Pallis, Marios D. Dikaiakos

Large Language Models (LLMs) have transformed natural language processing, but concerns have emerged about their susceptibility to ideological manipulation, particularly in politically sensitive areas. Prior work has foc…

The Impact of Ideological Discourses in RAG: A Case Study with COVID-19 Treatments

2026-03-16 · Elmira Salari, Maria Claudia Nunes Delfino, Hazem Amamou, José Victor de Souza 외 arxiv

This paper studies the impact of retrieved ideological texts on the outputs of large language models (LLMs). While interest in understanding ideology in LLMs has recently increased, little attention has been given to thi…

Mapping and Influencing the Political Ideology of Large Language Models using Synthetic Personas

2024-12-19 · Pietro Bernardelle, Leon Fröhling, Stefano Civelli, Riccardo Lunardi 외

The analysis of political biases in large language models (LLMs) has primarily examined these systems as single entities with fixed viewpoints. While various methods exist for measuring such biases, the impact of persona…

Influencing a Polarized and Connected Legislature

2022-05-16 · Ratul Das Chaudhury, C. Matthew Leister, Birendra Rai

When can an interest group exploit polarization between political parties to its advantage? Building upon Battaglini and Patacchini (2018), we study a model where an interest group credibly promises payments to legislato…

Political Ideology Shifts in Large Language Models

2025-08-22 · Pietro Bernardelle, Stefano Civelli, Leon Fröhling, Riccardo Lunardi 외 arxiv

Large language models (LLMs) are increasingly deployed in politically sensitive contexts, raising concerns about their susceptibility to ideological biases. In this work, we examine how synthetic persona conditioning sha…