Exploring Communication Strategies for Collaborative LLM Agents in Mathematical Problem-Solving

Large Language Model (LLM) agents are increasingly utilized in AI-aided education to support tutoring and learning. Effective communication strategies among LLM agents improve collaborative problem-solving efficiency and facilitate cost-effective adoption in education. However, little research has systematically evaluated the impact of different communication strategies on agents’ problem-solving. Our study examines four communication modes, $\mathit{\text{teacher-student}}\mathit{\text{interaction}}$, $\mathit{\text{peer-to-peer}}\mathit{\text{collaboration}}$, $\mathit{\text{reciprocal}}\mathit{\text{peer}}\mathit{\text{teaching}}$, and $\mathit{\text{critical}}\mathit{\text{debate}}$, in a dual-agent, chat-based mathematical problem-solving environment using the OpenAI GPT-4o model. Evaluated on the MATH dataset, our results show that dual-agent setups outperform single agents, with $\mathit{\text{peer-to-peer}}\mathit{\text{collaboration}}$ achieving the highest accuracy. Dialogue acts like statements, acknowledgment, and hints play a key role in collaborative problem-solving. While multi-agent frameworks enhance computational tasks, effective communication strategies are essential for tackling complex problems in AI education.