MLVU
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
MLVU
64.7Accuracy
20
MLVU (test)
62.75Accuracy
19
MLVU (651s)
78.1Accuracy
18
MLVU
68.4Base Accuracy
18
MLVU MCQ
82.1Accuracy
17
MLVU
75.1MLVU Score
16
MLVU (test)
75.7Accuracy
16
MLVU
73.83MLVU Score
16
MLVU
12.42MAT
16
MLVU 2024 (test)
54.9Accuracy
15
MLVU
64.66Accuracy
15
MLVU
80MLVU Score
14
MLVU long-form
74.7Accuracy
14
MLVU VS Eval
0.647Kendall's tau_b
13
MLVU
77.3Mean Score
12
MLVU multiple-choice questions
0.649Accuracy
12
MLVU
67.9p@1
11
MLVU 3–120 min
66.9Accuracy
11
MLVU
70.2M-Avg
11
MLVU Overall
53.8MLVU Score
10
MLVU (test)
73.8Accuracy
10
MLVU
77.8Performance Score
9
MLVU 2024 (test)
54.9Accuracy
9
MLVU
61,140Context Length (tokens)
8
MLVU
68.6M-Aug
8