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

Code Generation on BigCodeBench Lite-Pro Naive Stream

44.8Accuracy

ReAct

35.02437.56240.142.638Jun 1, 2026
Updated 1d ago

Evaluation Results

MethodLinks
2026.06
44.8--
2026.06
44.8--
2026.06
44.50.3-
2026.06
44.5-1.4
2026.06
43.81-
2026.06
43.8-2.1
2026.06
43.11.7-
2026.06
42.4-2.8
2026.06
42.4-1.4
2026.06
41.73.1-
2026.06
41-3.5
2026.06
39.65.2-
2026.06
39.6-4.2
2026.06
39.65.2-
2026.06
39.6-0
2026.06
35.49.4-