Ideology as a Problem: Lightweight Logit Steering for Annotator-Specific Alignment in Social Media Analysis

TL;DR

This paper presents a lightweight linear probe method to measure and correct ideological misalignment in large language models for social media analysis, without retraining the entire model.

Contribution

It introduces a simple, efficient bias correction technique that aligns model outputs with user-specific ideological preferences by adjusting output probabilities based on internal features.

Findings

Effective quantification of ideological misalignment

Minimal correction improves alignment with user opinions

Preserves model reasoning while adjusting outputs

Abstract

LLMs internally organize political ideology along low-dimensional structures that are partially, but not fully aligned with human ideological space. This misalignment is systematic, model specific, and measurable. We introduce a lightweight linear probe that both quantifies the misalignment and minimally corrects the output layer. This paper introduces a simple and efficient method for aligning models with specific user opinions. Instead of retraining the model, we calculated a bias score from its internal features and directly adjusted the final output probabilities. This solution is practical and low-cost and preserves the original reasoning power of the model.