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

Repository-level patch generation on SWE-Bench Lite (held-out 25% eval)

102.7Utility Difference (Δπ)

Rfx

7.95632.55357.1581.747Jun 23, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.06
102.7
2026.06
62
2026.06
62
2026.06
24.4
2026.06
21.7
2026.06
17.3
2026.06
13.7
2026.06
11.6