Exploration and Findings:

• The adoption of a human-in-the-loop approach using GPT-4-Turbo has shown promise in summarizing and categorizing synergistic learning during student interactions.
• Initial results suggest that this model could match human-level performance in understanding and facilitating collaborative discourse.

Implications for Future Research: This preliminary study indicates that further enhancements and refinements in the model could provide deeper insights into student collaborative behaviors and learning processes, paving the way for more effective educational models.