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

Zero-shot Evaluation on Zero-shot Benchmarks Average

68.6Average Accuracy

BF16

31.36841.03450.760.366Jun 9, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.06
68.6
2026.06
68.6
2026.06
68.3
2026.06
68.1
2026.06
67.4
2026.06
67.3
2026.06
67
2026.06
65.6
2026.06
65.2
2026.06
64.7
2026.06
63
2026.06
62.7
2026.06
62.6
2026.06
62
2026.06
62
2026.06
61.4
2026.06
61.2
2026.06
60.5
2026.06
60.4
2026.06
60.2
2026.06
60.1
2026.06
60
2026.06
59.9
2026.06
59.7
2026.06
59.4
2026.06
57.9
2026.06
55.9
2026.06
55.6
2026.06
55.6
2026.06
55.5
2026.06
55.1
2026.06
54.6
2026.06
53.5
2026.06
52.6
2026.06
52.6
2026.06
52.1
2026.06
52
2026.06
51.5
2026.06
51.4
2026.06
51.3
2026.06
51.1
2026.06
50.2
2026.06
50
2026.06
49.8
2026.06
49.3
2026.06
48.6
2026.06
48
2026.06
47.5
2026.06
47
2026.06
46.3
2026.06
46.3
2026.06
46.1
2026.06
44.2
2026.06
42.4
2026.06
40.6
2026.06
40.1
2026.06
39.6
2026.06
38
2026.06
35.7
2026.06
32.8