Prompt design impacts on-device LLM energy use, study finds

A new paper explores the impact of prompt design on the energy consumption of on-device Large Language Models (LLMs). Researchers conducted experiments on a smartphone, measuring power usage to understand how different prompt wordings affect energy efficiency. The study found that linguistic features, such as imperative keywords and instruction structure, significantly influence decoding length and overall energy consumption, suggesting prompt engineering as a viable method for optimizing on-device LLM performance. AI

IMPACT Prompt engineering offers a lightweight method to improve the energy efficiency of LLMs on resource-constrained devices.

RANK_REASON This is a research paper detailing an empirical study on LLM energy consumption. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.AI →

• alphaXiv
• arXiv
• CatalyzeX
• DagsHub
• Gotit.pub
• Hugging Face
• Influence Flower
• Large Language Models
• ScienceCast

• paper
• infra

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →