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

Natural Language Visual Reasoning on NLVR2 (test-p)

92.6Accuracy

BEIT-3

81.742484.561287.3890.1988Jan 2, 2021Sep 24, 2021Jun 17, 2022Mar 10, 2023Dec 1, 2023Aug 23, 2024May 16, 2025
Updated 1mo ago

Evaluation Results

MethodLinks
2023.05
92.6
2022.08
92.58
2021.11
89.54
2022.11
89.4
2023.05
88.3
2022.11
87.6
2022.05
87
2022.08
87
2023.05
87
2022.12
87
2022.11
87
2022.05
86.9
2022.11
86.9
2021.11
86.86
2022.06
86.86
2022.05
86.86
2022.11
86.1
2022.06
85.52
2022.05
85.2
2023.05
85.2
2021.08
85.15
2021.11
85.15
2022.08
85.15
2022.06
85.15
2022.05
84.95
2021.08
84.84
2021.11
84.84
2023.05
84.84
2022.05
84.84
2022.11
84.8
2021.11
84.76
2022.06
84.76
2023.05
84.76
2022.11
84.27
2021.11
84.21
2023.05
84.15
2022.05
84
2021.01
83.98
2021.08
83.98
2021.11
83.98
2021.11
83.98
2022.08
83.98
2022.12
83.98
2022.10
83.48
2022.06
83.47
2021.11
83.34
2022.06
83.34
2022.11
83.34
2023.05
83.34
2023.05
83.34
2022.10
83.34
2022.11
83.3
2022.06
83.22
2022.12
83.17
2021.07
83.14
2022.01
83.14
2021.11
83.14
2022.08
83.14
2022.11
83.14
2023.05
83.14
2023.05
83.14
2022.12
83.14
2022.05
83.14
2022.10
83.14
2023.12
83.14
2022.05
83.1
2023.04
83.1
2023.05
83.1
2022.11
83.1
2022.11
83.1
2023.05
83.09
2021.01
83.08
2021.02
83.08
2022.01
83.08
2022.02
83.08
2023.05
83.08
2022.06
83.08
2022.10
83.08
2021.11
83.05
2022.11
83.05
2023.05
83.05
2022.05
83.05
2022.10
83.05
2023.05
83
2022.11
82.7
2023.10
82.6
2021.11
82.47
2023.05
82.47
2022.01
82.3
2022.12
82.3
2022.05
82.3
2023.12
82.3
2025.05
82.28
2022.01
82.24
2022.08
82.24
2022.06
82.24
2022.11
82.24
2023.05
82.24
2022.05
82.2
2023.12
82.16
Showing 100 of 346 rows