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

Code Generation on BigCodeBench

83.84Accuracy

LatentMem

12.100830.725449.3567.9746Dec 4, 2025Dec 19, 2025Jan 3, 2026Jan 19, 2026Feb 3, 2026Feb 18, 2026Mar 6, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.02
83.84
2026.02
83.84
2026.02
83.25
2026.02
82.75
2026.02
82.67
2026.02
82.28
2026.02
82.28
2026.02
82.19
2026.02
82.11
2026.02
82.11
2026.02
82.11
2026.02
81.93
2026.02
81.93
2026.02
81.75
2026.02
81.67
2026.02
81.58
2026.02
81.49
2026.02
81.49
2026.02
81.46
2026.02
81.3
2026.02
81.28
2026.02
81.28
2026.02
81.11
2026.02
80.96
2026.02
80.28
2026.02
80.28
2026.02
80.26
2026.02
79.28
2026.03
78.16
2026.01
70.1
2026.03
65.92
2026.03
64.12
2026.03
63.51
2026.01
63
2026.01
62.9
2026.01
61.7
2026.03
53.3
2026.03
51.7
2026.03
51.67
2026.03
50.7
2026.03
50.1
2026.03
45.3
2025.12
41.9
2025.12
41.4
2025.12
41.1
2025.12
37.86
37.51
2025.12
37.5
37.16
2025.12
37
2025.12
36.9
2025.12
36.8
2025.12
36.2
2026.01
35.9
2025.12
35.5
2025.12
35.3
2026.01
35.2
2026.01
33.8
2025.12
33.3
2026.01
33.1
33.07
2025.12
32.75
2026.01
32.4
2026.01
32.4
2026.01
32.4
2025.12
31.9
2025.12
30.41
2026.03
29.91
2026.03
28.42
2025.12
28.38
2025.12
14.86