List your company

Correctness simulation accuracy on DBE-KT22 50% historical interaction logs

71.27AccuracyEdu-Theater
62.9565.1167.2769.43Jun 13, 2026

Evaluation results

MethodLinks
Edu-TheaterBase LLM=GPT-4.1-mini2026.0671.27
Agent4EduBase LLM=GPT-4.1-mini2026.0669.95
Edu-TheaterBase LLM=Llama3-8b2026.0669.67
Agent4EduBase LLM=Llama3-8b2026.0669.63
EERNN2026.0669.04
EduAgentBase LLM=GPT-4.1-mini2026.0668.52
DAISIM2026.0668.44
SAKT2026.0668.37
EduAgentBase LLM=Llama3-8b2026.0667.33
DKVMN2026.0665.33
KES2026.0663.27