Scaling Laws for Educational AI Agents

2026-03-12

AI Summary

This work challenges the prevailing paradigm that relies solely on scaling large language models to enhance educational AI agents, proposing instead a structured approach grounded in an Agent Scaling Law. It introduces EduClaw, a multi-agent platform centered on the AgentProfile mechanism, which systematically expands agent capabilities along five dimensions: role definition, skill depth, tool completeness, runtime adaptability, and infusion of educator knowledge. Leveraging a JSON-based specification, modular skill design, and a dual-axis expansion strategy coupling tools with skills, the platform successfully deploys over 330 agent configurations encompassing more than 1,100 skill modules across K–12 education. Empirical results demonstrate that agent performance predictably improves with structural richness, offering the first systematic characterization of capability growth dynamics in educational AI agents.

Technology Category

• Multiagent Systems: Modeling other Agents
• Cognitive Modeling & Cognitive Systems: Agent Architectures
• Humans and AI: Human-Aware Planning and Behavior Prediction

Application Category

• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Search and Retrieval-Augmented AI: Agentic search
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

While scaling laws for Large Language Models (LLMs) have been extensively studied along dimensions of model parameters, training data, and compute, the scaling behavior of LLM-based educational agents remains unexplored. We propose that educational agent capability scales not merely with the underlying model size, but through structured dimensions that we collectively term the Agent Scaling Law: role definition clarity, skill depth, tool completeness, runtime capability, and educator expertise injection. Central to this framework is AgentProfile, a structured JSON-based specification that serves as the mechanism enabling systematic capability growth of educational agents. We present EduClaw, a profile-driven multi-agent platform that operationalizes this scaling law, demonstrating its effectiveness through the construction and deployment of 330+ educational agent profiles encompassing 1,100+ skill modules across K-12 subjects. Our empirical observations suggest that educational agent performance scales predictably with profile structural richness. We identify two complementary scaling axes -- Tool Scaling and Skill Scaling -- as future directions, arguing that the path to more capable educational AI lies not solely in larger models, but in stronger structured capability systems.

Problem

• Scaling Laws
• Educational AI Agents
• Agent Capability
• Structured Specification
• Profile-based Scaling

Innovation

• Agent Scaling Law
• AgentProfile
• EduClaw
• Skill Scaling
• Tool Scaling