Small Language Models for Curriculum-based Guidance

2025-09-27

AI Summary

This study addresses the high energy consumption and deployment challenges of large language models (LLMs) in educational settings by proposing a lightweight, sustainable, curriculum-aware AI teaching assistant framework. Methodologically, it integrates retrieval-augmented generation (RAG), eight open-source small language models (SLMs)—including LLaMA 3.1, Granite 3.3, and Gemma 3—and pedagogy-oriented prompt engineering to enable real-time, curriculum-knowledge-grounded responses. Its core contributions are threefold: (1) the first systematic empirical validation that optimized SLMs can achieve teaching alignment and response accuracy comparable to GPT-4o, with an average performance gap of less than 3.2%; (2) full offline execution on consumer-grade GPUs; and (3) drastic efficiency gains—87% reduction in inference energy consumption and 92% lower computational requirements—thereby enabling privacy-preserving, green AI deployment in resource-constrained educational environments.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Natural Language Processing: (Large) Language Models
• Search and Optimization: Learning to Search

Application Category

• Search and Retrieval-Augmented AI: Search Tool Learning with LLM: Teaching LLMs to invoke search and make use of retrieved information
• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation
• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems

Abstract

The adoption of generative AI and large language models (LLMs) in education is still emerging. In this study, we explore the development and evaluation of AI teaching assistants that provide curriculum-based guidance using a retrieval-augmented generation (RAG) pipeline applied to selected open-source small language models (SLMs). We benchmarked eight SLMs, including LLaMA 3.1, IBM Granite 3.3, and Gemma 3 (7-17B parameters), against GPT-4o. Our findings show that with proper prompting and targeted retrieval, SLMs can match LLMs in delivering accurate, pedagogically aligned responses. Importantly, SLMs offer significant sustainability benefits due to their lower computational and energy requirements, enabling real-time use on consumer-grade hardware without depending on cloud infrastructure. This makes them not only cost-effective and privacy-preserving but also environmentally responsible, positioning them as viable AI teaching assistants for educational institutions aiming to scale personalized learning in a sustainable and energy-efficient manner.

Problem

• Developing AI teaching assistants using small language models
• Providing curriculum-based guidance through retrieval-augmented generation
• Enabling sustainable educational AI with lower computational requirements

Innovation

• RAG pipeline with small language models
• SLMs match LLMs via targeted retrieval
• Energy-efficient real-time use on consumer hardware