swe-bench-verified · run 2481

Score delta by case

candidate 2.4.1-rc3 vs baseline 2.3.0

26 cases · 4 improved · 4 regressed · 18 unchanged

swe-bench-verified cases, sorted by score delta
CaseBaseCandΔ scoreΔ
astropy-129070.410.60+0.190
sympy-205900.340.47+0.130
django-110390.550.63+0.080
pytest-73730.720.75+0.030
sympy-169880.440.44+0.003
requests-23170.810.81+0.002
seaborn-30100.630.63+0.002
django-157380.740.74+0.002
matplotlib-253110.570.57+0.002
matplotlib-239130.680.680.000
sphinx-103250.520.520.000
flask-50630.900.900.000
django-134010.580.580.000
astropy-143650.710.710.000
psf-black-23130.660.660.000
scikit-257470.500.500.000
pytest-111480.430.430.000
requests-19630.860.860.000
pylint-65060.470.47−0.001
mwaskom-31870.360.36−0.001
sympy-181890.610.61−0.002
xarray-33640.390.39−0.002
xarray-40940.490.45−0.040
sphinx-87210.380.29−0.090
django-168190.600.47−0.130
scikit-134970.440.29−0.150
Δ score · candidate − baseline0.200+0.20

Bars right of zero are gains, bars left are regressions.