LLMs simulate policy responses with new transaction-cost persona method

Researchers have developed a new method for simulating how individuals might respond to energy policy interventions by incorporating the concept of perceived transaction costs into Large Language Model (LLM) personas. This approach, tested using survey data from the Netherlands, represents tenants not just by their demographics but by their understanding of the costs, benefits, and barriers associated with energy-efficient renovations. The study found that this friction-aware persona design consistently improved model performance across various LLMs, including GPT-3.5 Turbo, Ministral-8B-Instruct, and Llama 3.1 8B-Instruct, suggesting its utility for policy simulation. AI

IMPACT Enhances LLM capabilities for simulating complex human behaviors in policy contexts.

RANK_REASON Academic paper detailing a new methodology for LLM simulation. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.AI →

• Energy-efficient renovation
• GPT-3.5 Turbo
• Group Relative Policy Optimization
• Llama 3.1 8B-Instruct
• Ministral-8B-Instruct
• Netherlands
• Perceived transaction cost
• supervised fine-tuning

• paper
• policy
• other

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