Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Quality-judgment accuracy on LLMBar ZH condition

86.5Strict Accuracy (%)

DeepSeek

71.10875.10479.183.096Jun 12, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.06
86.587.885.653.12.5
2026.06
83.383.781.256.90.8
2026.06
7474.371.4530.7
2026.06
71.772.869.852.62.1