Legal Translation on HKCFA Judgment 97-22 benchmark leaderboard · SOTA2 Research

Legal Translation on HKCFA Judgment 97-22

85.12xCOMET-XLGPT-4o

Evaluation Results

Method	Links
GPT-4o2025.07		85.12	91.78	88.45
GPT-42025.07		84.24	90.1	87.15
MoE-16B2025.07		83.53	89.56	86.55
R1-Distill-32B2025.07		83.25	89.33	86.29
ChatGPT2025.07		82.29	88.52	85.41
14B-Chat2025.07		81.86	87.12	84.49
13B-Chat2025.07		81.33	87.51	84.42
7B-Chat2025.07		80.54	87.06	83.8
3-6B2025.07		80.27	87.2	83.74
13B-Base2025.07		79.6	86.51	83.06
33B2025.07		79.29	85.8	82.55
2-6B2025.07		78.11	84.78	81.45
13B2025.07		76.26	82.59	79.43
GPT-4o2025.07		69.42	75.88	72.65
GPT-42025.07		68.1	74.25	71.17
MoE-16B2025.07		67.45	73.45	70.45
R1-Distill-32B2025.07		67.12	73.12	70.12
ChatGPT2025.07		66.15	72.1	69.12
14B-Chat2025.07		65.55	71.8	68.67
13B-Chat2025.07		64.4	70.5	67.45
7B-Chat2025.07		63.9	70.15	67.02
3-6B2025.07		63.85	69.2	66.52
33B2025.07		62.5	68.8	65.65
13B-Base2025.07		62.2	69.1	65.65
2-6B2025.07		61.1	67.5	64.3
13B2025.07		58.5	64.8	61.65