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

Long-context language generation on RepoBench-P

4.46Average Acceptance Length

LongSpec

0.86161.79582.733.6642Feb 24, 2025
Updated 11d ago

Evaluation Results

MethodLinks
2025.02
4.4696.963.26
2025.02
4.03115.272.7
2025.02
3.86110.762.38
2025.02
3.5977.222.65
2025.02
3.3991.281.69
2025.02
2.9436.661.26
2025.02
2.8535.671.2
2025.02
2.7189.222.09
2025.02
2.6144.390.82
2025.02
2.5748.751.05
2025.02
2.5744.131.03
2025.02
2.2374.151.59
2025.02
1.841.071.38
2025.02
1.3845.70.85
2025.02
1.3230.61.05
2025.02
119.18-
2025.02
146.611
2025.02
113.44-
2025.02
129.141
2025.02
117.02-
2025.02
142.691
2025.02
113.85-
2025.02
129.741
2025.02
122.77-
2025.02
154.081