SuperCLUE专业技能与知识排行榜(2023年9月)
排名 |
模型 |
总分 |
计算 |
逻辑与推理 |
代码 |
知识与百科 |
1 |
GPT4 |
94.9 |
95.56 |
100 |
85.89 |
98.14 |
2 |
Claude2 |
84.56 |
75.48 |
100 |
74.63 |
88.14 |
3 |
GPT3.5 |
79.49 |
74.04 |
95.1 |
69.25 |
79.56 |
4 |
Moonshot |
72.92 |
64.81 |
100 |
44.74 |
82.14 |
5 |
SenseChat 3.0 |
69.79 |
43.4 |
88.16 |
58.57 |
89.02 |
6 |
ChatGLM2-Pro |
67.26 |
64.81 |
90.54 |
36.84 |
76.83 |
7 |
云雀大模型(豆包) |
63.07 |
43.52 |
93.42 |
26.32 |
89.02 |
8 |
MiniMax-Abab5.5 |
56.89 |
34.26 |
63.51 |
47.37 |
82.43 |
9 |
Baichuan2-13B-Chat |
56.81 |
50.93 |
80.26 |
36.84 |
59.21 |
10 |
通义千问plus |
55.34 |
46.3 |
70 |
35.53 |
69.51 |
11 |
讯飞星火V2.0 |
54.53 |
51.85 |
55.41 |
31.58 |
79.27 |
12 |
OpenBuddy-70B |
54.51 |
31.48 |
89.19 |
47.37 |
50 |
13 |
Qwen-14B-Chat |
53.94 |
52.78 |
52.86 |
44.74 |
65.38 |
14 |
Chinese_Alpaca_2_13B |
39.9 |
24.07 |
52.7 |
35.53 |
47.3 |
15 |
360GPT_S2_V9 |
32.4 |
13.89 |
64.86 |
16.22 |
34.62 |
16 |
ChatGLM2-6B |
31.66 |
18.52 |
58.11 |
25 |
25 |
17 |
Llama2-13B-Chat |
26.17 |
7.41 |
48.53 |
32.89 |
15.85 |