Correctness Simulation Accuracy on DBE-KT22 (0% historical interaction logs)
68.85AccuracyEdu-Theater
Evaluation Results
| Method | Links | |
|---|---|---|
| Edu-TheaterBase LLM=GPT-4.1-mini2026.06 | 68.85 | |
| Agent4EduBase LLM=GPT-4.1-mini2026.06 | 68.43 | |
| EduAgentBase LLM=GPT-4.1-mini2026.06 | 66.93 | |
| Edu-TheaterBase LLM=Llama3-8b2026.06 | 66.5 | |
| Agent4EduBase LLM=Llama3-8b2026.06 | 65.83 | |
| EduAgentBase LLM=Llama3-8b2026.06 | 63.25 |