In the rapidly evolving field of artificial intelligence, large language models (LLMs) are being applied to agriculture, specifically in pest management. This study evaluates the feasibility of using LLMs like GPT-4 to generate pest management advice. The evaluation focuses on various aspects such as Coherence, Logical Consistency, Fluency, Relevance, Comprehensibility, and Exhaustiveness. Results show that GPT-3.5 and GPT-4 outperform FLAN models in most categories. Instruction-based prompting with domain-specific knowledge significantly improves accuracy rates up to 72%. Different prompting methods impact the linguistic quality and performance metrics of the models differently. Overall, LLMs show promise but require continuous updating and fine-tuning for specific domains like agriculture.
To Another Language
from source content
arxiv.org
Principais Insights Extraídos De
by Shanglong Ya... às arxiv.org 03-19-2024
https://arxiv.org/pdf/2403.11858.pdfPerguntas Mais Profundas