Inference-Time Mitigation of Adversarial Political Bias in Large Language Models

• URL: http://arxiv.org/abs/2608.14629v1
• Date: Fri, 24 Jul 2026 01:52:10 GMT
• Title: Inference-Time Mitigation of Adversarial Political Bias in Large Language Models
• Abstract summary: We present different methods to shield Large Language Models (LLMs) against the injection of political bias.<n>Our results demonstrate that the proposed Recursive Self-Correction approach raises model performance from a Political Neutrality Likert scale baseline of 2.14 to 4.56.
• Score: 11.657742684742258
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: As Large Language Models (LLMs) become the mainstay for information retrieval and summarization tasks, ensuring that they are always non-partisan and invulnerable to political bias is a critical step towards safer and more trustworthy Artificial Intelligence (AI). Current model alignment paradigms, such as reinforcement learning from human feedback (RLHF), make LLMs follow overarching safety instructions. However, this instruction tuning can be exploited via adversarial prompt injection and be used to generate unsafe content. In particular, political bias has not been specifically targeted by modern alignment techniques as harmful and biased content. To address this vulnerability of LLMs, we propose mitigation strategies using Chain of Thought (CoT) prompting and Direct Preference Optimization (DPO). Using a public dataset of legislative videos, we generate summaries using LLMs, inject bias via adversarial prompting and evaluate their performance on a four axis scale designed for political summarization. In this paper, we present different methods to shield LLMs against the injection of political bias. Our results demonstrate that the proposed Recursive Self-Correction approach raises model performance from a Political Neutrality Likert scale baseline of 2.14 to 4.56, averaged across all models, demonstrating effective inference-time mitigation of political bias in LLM-generated summaries.

Related papers

• A Multi-Dimensional Audit of Politically Aligned Large Language Models [3.147897243459717]
  This work aims to ensure politically-aligned language models generate legitimate, harmless arguments.<n>We propose a multi-dimensional framework inspired by Habermas' Theory of Communicative Action to audit politically aligned language models.
  arXiv  Detail & Related papers  (2026-04-27T12:57:52Z)
• Policy-Conditioned Policies for Multi-Agent Task Solving [53.67744322553693]
  In this work, we propose a paradigm shift that bridges the gap by representing policies as human-interpretable source code.<n>We reformulate the learning problem by utilizing Large Language Models (LLMs) as approximate interpreters.<n>We formalize this process as textitProgrammatic Iterated Best Response (PIBR), an algorithm where the policy code is optimized by textual gradients.
  arXiv  Detail & Related papers  (2025-12-24T07:42:10Z)
• Steering Towards Fairness: Mitigating Political Bias in LLMs [16.594400974742523]
  We employ a framework for probing and mitigating such biases in large language models (LLMs) through analysis of internal model representations.<n>We introduce a comprehensive activation extraction pipeline capable of layer-wise analysis across multiple ideological axes.<n>Our results show that decoder LLMs systematically encode representational bias across layers, which can be leveraged for effective steering vector-based mitigation.
  arXiv  Detail & Related papers  (2025-08-12T11:09:03Z)
• PRISM: A Methodology for Auditing Biases in Large Language Models [9.751718230639376]
  PRISM is a flexible, inquiry-based methodology for auditing Large Language Models. It seeks to illicit such positions indirectly through task-based inquiry prompting rather than direct inquiry of said preferences.
  arXiv  Detail & Related papers  (2024-10-24T16:57:20Z)
• Evaluating Implicit Bias in Large Language Models by Attacking From a Psychometric Perspective [66.34066553400108]
  We conduct a rigorous evaluation of large language models' implicit bias towards certain demographics.<n>Inspired by psychometric principles, we propose three attack approaches, i.e., Disguise, Deception, and Teaching.<n>Our methods can elicit LLMs' inner bias more effectively than competitive baselines.
  arXiv  Detail & Related papers  (2024-06-20T06:42:08Z)
• Fine-Tuning Language Models with Reward Learning on Policy [68.70065254564642]
  Reinforcement learning from human feedback (RLHF) has emerged as an effective approach to aligning large language models (LLMs) to human preferences. Despite its popularity, (fixed) reward models may suffer from inaccurate off-distribution. We propose reward learning on policy (RLP), an unsupervised framework that refines a reward model using policy samples to keep it on-distribution.
  arXiv  Detail & Related papers  (2024-03-28T10:02:10Z)
• Whose Side Are You On? Investigating the Political Stance of Large Language Models [56.883423489203786]
  We investigate the political orientation of Large Language Models (LLMs) across a spectrum of eight polarizing topics. Our investigation delves into the political alignment of LLMs across a spectrum of eight polarizing topics, spanning from abortion to LGBTQ issues. The findings suggest that users should be mindful when crafting queries, and exercise caution in selecting neutral prompt language.
  arXiv  Detail & Related papers  (2024-03-15T04:02:24Z)
• Nash Learning from Human Feedback [86.09617990412941]
  We introduce an alternative pipeline for the fine-tuning of large language models using pairwise human feedback. We term this approach Nash learning from human feedback (NLHF) We present a novel algorithmic solution, Nash-MD, founded on the principles of mirror descent.
  arXiv  Detail & Related papers  (2023-12-01T19:26:23Z)
• 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)

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.