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

Reading Comprehension on DROP (accuracy)

88.8DROP Accuracy

Direct Fine-tuning

-0.43222.73445.969.066Oct 20, 2022May 15, 2023Dec 8, 2023Jul 3, 2024Jan 26, 2025Aug 21, 2025Mar 17, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.02
88.8
2026.02
88.66
2026.02
88.25
2026.02
88.03
2026.02
87.97
2026.02
87.7
2026.02
87.67
2026.02
86.41
2022.10
84.9
2022.10
83
2026.02
78.24
2022.10
78.2
2026.02
78.17
2026.02
76.58
2026.02
76.39
2022.10
76.2
2026.02
76.18
2026.02
75.72
2026.02
75.21
2025.12
73
2022.10
71.7
2025.12
71.2
2022.10
70.6
2026.02
69.83
2025.12
65.2
2022.10
60
2025.12
60
2025.12
58.2
2026.02
56.7
2026.02
56.5
2026.02
53.1
2026.02
52.6
2026.02
52.4
2026.02
47.8
2026.02
43.8
2025.12
28
2026.03
20.5
2026.03
19.6
2026.03
19.5
2026.03
19.5
2026.03
19.4
2026.03
18.7
2026.03
18.4
2025.12
17
2025.12
17
2025.12
17
2026.03
16.8
2025.12
16
2025.12
16
2025.12
16
2025.12
15
2025.12
15
2025.12
15
2025.12
14
2025.12
14
2025.12
14
2025.12
14
2025.12
13
2025.12
13
2025.12
13
2025.12
13
2025.12
12
2025.12
12
2025.12
12
2025.12
12
2025.12
12
2025.12
11
2025.12
11
2025.12
11
2025.12
11
2025.12
11
2025.12
11
2025.12
11
2025.12
11
2025.12
10
2025.12
10
2025.12
10
2025.12
10
2025.12
9
2025.12
9
2025.12
8
2025.12
8
2025.12
8
2025.12
7
2025.12
7
2025.12
6
2025.12
6
2025.12
6
2025.12
6
2025.12
5
2025.12
5
2025.12
5
2025.12
4
2025.12
4
2025.12
4
2025.12
4
2025.12
4
2025.12
3
2025.12
3
2025.12
3
Showing 100 of 111 rows