Revisiting OPRO: The Limitations of Small-Scale LLMs as Optimizers
This work highlights a practical limitation for researchers and practitioners using small-scale LLMs in automated prompting, suggesting it is incremental by focusing on model-specific constraints rather than a new paradigm.
The paper investigates the effectiveness of the OPRO method for automated prompting using small-scale LLMs like LLaMa-2 and Mistral 7B, finding it limited due to their constrained inference capabilities, and recommends direct instructions as robust baselines for efficient prompt engineering.
Numerous recent works aim to enhance the efficacy of Large Language Models (LLMs) through strategic prompting. In particular, the Optimization by PROmpting (OPRO) approach provides state-of-the-art performance by leveraging LLMs as optimizers where the optimization task is to find instructions that maximize the task accuracy. In this paper, we revisit OPRO for automated prompting with relatively small-scale LLMs, such as LLaMa-2 family and Mistral 7B. Our investigation reveals that OPRO shows limited effectiveness in small-scale LLMs, with limited inference capabilities constraining optimization ability. We suggest future automatic prompting engineering to consider both model capabilities and computational costs. Additionally, for small-scale LLMs, we recommend direct instructions that clearly outline objectives and methodologies as robust prompt baselines, ensuring efficient and effective prompt engineering in ongoing research.