اختبار مرتبط بالدليل · متوسط المستوى

SWE-bench and Coding Agent Benchmarks Quiz

Tests how SWE-bench tasks are built, FAIL_TO_PASS and PASS_TO_PASS scoring, the Lite and Verified subsets, and why scaffolds change results.

مسارات الدليل ذات الصلةSwe Bench And Coding Agent Benchmarks

سؤال 1 من 8

What does a SWE-bench task ask the system to do?