EduThink4AI: Translating Educational Critical Thinking into Multi-Agent LLM Systems

2025-07-20

AI Summary

Existing LLM-based educational systems exhibit significant limitations in fostering critical thinking: they struggle with multi-hop reasoning under counterfactual premises and are vulnerable to adversarial prompts, leading to bias and factual inaccuracies. To address this, we propose EDU-Prompting—the first framework that systematically integrates pedagogical theories of critical thinking into a multi-agent collaborative pipeline. It employs modular prompt design, role-assigned agent coordination, and an interpretable reasoning mechanism to generate de-biased, multi-perspective, and logically rigorous instructional responses. Comprehensive evaluation on theoretical benchmarks and university-level writing tasks demonstrates that EDU-Prompting substantially improves response veracity and inferential rigor. It outperforms mainstream baselines on multi-hop counterfactual reasoning and critical writing tasks, while maintaining strong compatibility with existing LLM-based educational infrastructures.

Technology Category

• Natural Language Processing: Prompt Engineering / Prompting
• Machine Learning: Large Multimodal Models (LMMs)
• Multiagent Systems: Adversarial Agents

Application Category

• Search and Retrieval-Augmented AI: Search Tool Learning with LLM: Teaching LLMs to invoke search and make use of retrieved information
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems

Abstract

Large language models (LLMs) have demonstrated significant potential as educational tutoring agents, capable of tailoring hints, orchestrating lessons, and grading with near-human finesse across various academic domains. However, current LLM-based educational systems exhibit critical limitations in promoting genuine critical thinking, failing on over one-third of multi-hop questions with counterfactual premises, and remaining vulnerable to adversarial prompts that trigger biased or factually incorrect responses. To address these gaps, we propose EDU-Prompting, a novel multi-agent framework that bridges established educational critical thinking theories with LLM agent design to generate critical, bias-aware explanations while fostering diverse perspectives. Our systematic evaluation across theoretical benchmarks and practical college-level critical writing scenarios demonstrates that EDU-Prompting significantly enhances both content truthfulness and logical soundness in AI-generated educational responses. The framework's modular design enables seamless integration into existing prompting frameworks and educational applications, allowing practitioners to directly incorporate critical thinking catalysts that promote analytical reasoning and introduce multiple perspectives without requiring extensive system modifications.

Problem

• Enhancing critical thinking in LLM-based educational tutoring systems
• Reducing bias and factual errors in AI-generated educational responses
• Integrating educational theories into multi-agent LLM frameworks

Innovation

• Multi-agent LLM framework for critical thinking
• EDU-Prompting integrates educational theories
• Modular design enhances truthfulness and logic