Rethinking Scale: The Efficacy of Fine-Tuned Open-Source LLMs in Large-Scale Reproducible Social Science Research

Short summary

Fine-tuned open-source LLMs, such as Llama2-7B, achieve performance comparable to large closed-source models like GPT-4 on social science text classification tasks, making them a reproducible and low-cost alternative for large-scale research.

AI-generated from the title and abstract; the full text is not read.