Our new X account is live! Follow @wizwand_team for updates

LLM-as-a-Judge on PreferenceBench

0.69Rstd

CalibraEval

Updated 4d ago

Evaluation Results

Method	Links
CalibraEval 2024.10		0.69	90.71
DI 2024.10		0.8	90.21
Pride 2024.10		2.27	90.26
CalibraEval 2024.10		2.42	83.98
Qwen-72B 2024.10		2.54	90.12
CalibraEval 2024.10		2.82	85.98
DC 2024.10		2.9	83.65
ChatGPT 2024.10		3.04	85.61
CC 2024.10		3.13	84.83
Llama-3-8B 2024.10		3.36	83.43
Pride 2024.10		3.51	85.68
DI 2024.10		4.03	83.45
CalibraEval 2024.10		4.29	90.49
Pride 2024.10		4.35	83.24
DI 2024.10		4.49	90.46
CalibraEval 2024.10		5.12	83.88
Pride 2024.10		5.24	90.47
GPT4o 2024.10		5.29	90.44
CC 2024.10		5.41	89.3
CC 2024.10		5.79	90.02
DC 2024.10		5.81	82.68
DC 2024.10		5.91	77.98
DC 2024.10		5.99	88.22
CalibraEval 2024.10		6.19	78.64
CC 2024.10		6.31	77.91
Llama-3.1-8B 2024.10		6.65	77.54
CC 2024.10		7.03	78.41
DC 2024.10		7.04	88.83
Pride 2024.10		7.36	83.55
CC 2024.10		9.47	82.78
Pride 2024.10		9.47	77.92
DI 2024.10		9.82	83.94
DC 2024.10		10.07	79.96
Qwen-14B 2024.10		11.99	80.68
DI 2024.10		12.42	75.14
DI 2024.10		15.77	76.14