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

Science Question Answering on ARC Challenge (Character-level Accuracy)

76.1Character-level Accuracy

Full

21.91635.98350.0564.117Mar 16, 2026Mar 29, 2026Apr 11, 2026Apr 24, 2026May 7, 2026May 20, 2026Jun 3, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.06
76.1
2026.06
75.1
2026.06
74.9
2026.06
73
2026.06
70.3
2026.06
68.3
2026.06
66.9
2026.06
66.8
2026.06
66.7
2026.06
66
2026.06
65.7
2026.06
64.5
2026.06
62.7
2026.06
61.8
2026.06
57.5
2026.06
57.1
2026.03
52.2
2026.03
50.56
2026.03
49.4
2026.06
48.4
2026.06
44.8
2026.03
44.54
2026.06
44.1
2026.06
43.2
2026.03
42.58
2026.03
41.47
2026.03
38.89
2026.03
38.57
2026.03
38.4
2026.03
38.28
2026.06
31.7
2026.06
29.6
2026.03
29
2026.06
27.8
2026.06
25.7
2026.06
24.2
2026.06
24