![[2602.16735] A Few-Shot LLM Framework for Extreme Day Classification in Electricity Markets](https://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png){kind=logo}
[2602.16735] A Few-Shot LLM Framework for Extreme Day Classification in Electricity Markets
Summary
This paper presents a few-shot classification framework utilizing Large Language Models (LLMs) to predict spikes in electricity prices, demonstrating comparable performance to traditional models with limited data.
Why It Matters
As electricity markets face increasing volatility, effective prediction of price spikes is crucial for market participants. This research highlights the potential of LLMs to provide data-efficient solutions, which can enhance decision-making in energy management and trading.
Key Takeaways
- The proposed framework uses LLMs for few-shot classification of electricity price spikes.
- It aggregates various system state information into natural-language prompts for LLM input.
- The model shows performance on par with supervised methods like SVM and XGBoost, especially with limited data.
- This approach highlights the versatility of LLMs in applications beyond traditional NLP tasks.
- The findings could lead to improved strategies for managing electricity market risks.
Computer Science > Machine Learning arXiv:2602.16735 (cs) [Submitted on 17 Feb 2026] Title:A Few-Shot LLM Framework for Extreme Day Classification in Electricity Markets Authors:Saud Alghumayjan, Ming Yi, Bolun Xu View a PDF of the paper titled A Few-Shot LLM Framework for Extreme Day Classification in Electricity Markets, by Saud Alghumayjan and 2 other authors View PDF HTML (experimental) Abstract:This paper proposes a few-shot classification framework based on Large Language Models (LLMs) to predict whether the next day will have spikes in real-time electricity prices. The approach aggregates system state information, including electricity demand, renewable generation, weather forecasts, and recent electricity prices, into a set of statistical features that are formatted as natural-language prompts and fed to an LLM along with general instructions. The model then determines the likelihood that the next day would be a spike day and reports a confidence score. Using historical data from the Texas electricity market, we demonstrate that this few-shot approach achieves performance comparable to supervised machine learning models, such as Support Vector Machines and XGBoost, and outperforms the latter two when limited historical data are available. These findings highlight the potential of LLMs as a data-efficient tool for classifying electricity price spikes in settings with scarce data. Subjects: Machine Learning (cs.LG); Systems and Control (eess.SY) Cite as: arXiv:2602.1...
