List your company

Educational Feedback Generation on SEFL (test)

57Win Rate (vs GPT-4o)SEFL
42.4446.225053.78Feb 18, 2025

Evaluation results

MethodLinks
SEFLBackbone=Qwen2.5-14B, Evaluation Protocol=LLM-as-a-Judge2025.02576373
Book2DialBackbone=Qwen2.5-14B, Evaluation Protocol=LLM-as-a-Judge, Data size=5,300 conversation pairs2025.02433727