MarxistLLM: Fine-tuning a language model with a Marxist worldview

Theoretical perspectives, such as rational choice theory, Marxist theory, and feminist theory, provide a defined worldview (Weltanschauung) that guides researchers' analytical and interpretive processes.Recently, there has been an increasing interest in using language models to examine social aspects through social science lenses.However, critical computing scholars have shown that models embody specific societal values and perspectives.This inherent encoding of viewpoints means that when these models are employed in research, scholars might not reflect the worldview or use theoretical perspectives.To address this issue, our study explores the possibility of fine-tuning large language models with a specific Weltanschauung. Specifically, we incorporate the writings of Marx and Engels to fine-tune these models, aiming to infuse them with Marxist ideological terminology and worldview.We evaluate how these fine-tuned models differ in empirical analysis.This investigation underscores the importance of aligning theoretical perspectives with computational tools to enhance the social sciences.