swebench-pro: by models

Home Paper Code


SE predicted by accuracy

The typical standard errors between pairs of models on this dataset as a function of the absolute accuracy.

SE vs accuracy difference

The standard error of each model pair against their observed accuracy difference. Pairs below a reference line differ by more than the corresponding number of standard errors, i.e. they are statistically distinguishable at that level.

CDF of question level accuracy

Results table by model

model pass1 pass@count win_rate count SE(A) SE_x(A) SE_pred(A)
cwm-baseline-400 31.1 31.1 10.7 1 1.7 NaN NaN
cwm-baseline-380 30.2 30.2 9.98 1 1.7 NaN NaN
cwm-baseline-210 29.5 29.5 9.73 1 1.7 NaN NaN
cwm-selfplay-280 29.5 29.5 9.88 1 1.7 NaN NaN
cwm-baseline_360-409 29.5 45.4 9.47 5 1.7 1.3 1
cwm-baseline-370 29.4 29.4 9.16 1 1.7 NaN NaN
cwm-baseline-340 29.3 29.3 9.52 1 1.7 NaN NaN
cwm-selfplay-120 29 29 9.43 1 1.7 NaN NaN
cwm-baseline_340-389 29 45.1 9.11 5 1.7 1.3 1
cwm-baseline-220 28.9 28.9 9.24 1 1.7 NaN NaN
cwm-selfplay-150 28.9 28.9 9.13 1 1.7 NaN NaN
cwm-baseline_350-399 28.8 44.3 9 5 1.7 1.3 1
cwm-selfplay-270 28.7 28.7 9.43 1 1.7 NaN NaN
cwm-selfplay-220 28.6 28.6 8.61 1 1.7 NaN NaN
cwm-baseline-270 28.6 28.6 8.68 1 1.7 NaN NaN
cwm-selfplay_120-169 28.6 45.6 8.99 5 1.7 1.3 1
cwm-baseline_330-379 28.5 44.2 8.79 5 1.7 1.3 1
cwm-selfplay-140 28.5 28.5 9.09 1 1.7 NaN NaN
cwm-selfplay-160 28.5 28.5 8.68 1 1.7 NaN NaN
cwm-baseline-390 28.5 28.5 8.96 1 1.7 NaN NaN
cwm-selfplay-260 28.5 28.5 8.81 1 1.7 NaN NaN
cwm-selfplay_240-289 28.4 44.6 9.08 4 1.7 1.3 1.1
cwm-selfplay-180 28.3 28.3 8.54 1 1.7 NaN NaN
cwm-baseline-320 28.3 28.3 8.92 1 1.7 NaN NaN
cwm-baseline-170 28.3 28.3 8.64 1 1.7 NaN NaN
cwm-selfplay-80 28.3 28.3 8.48 1 1.7 NaN NaN
cwm-selfplay_140-189 28.3 44.9 8.69 5 1.7 1.3 1
cwm-baseline_320-369 28.3 44 8.74 5 1.7 1.3 1
cwm-baseline_210-259 28.2 45.4 8.85 5 1.7 1.3 1.1
cwm-selfplay_130-179 28.2 45.3 8.75 5 1.7 1.3 1
cwm-baseline-360 28.2 28.2 8.76 1 1.7 NaN NaN
cwm-selfplay_110-159 28.1 44.9 8.75 5 1.7 1.3 1.1
cwm-baseline_310-359 28.1 44 8.62 5 1.7 1.3 1
cwm-baseline-130 28 28 8.11 1 1.7 NaN NaN
cwm-selfplay-130 28 28 8.82 1 1.7 NaN NaN
cwm-selfplay_150-199 28 44.2 8.53 5 1.7 1.3 1
cwm-baseline_200-249 28 45.3 8.67 5 1.7 1.3 1.1
cwm-baseline_190-239 28 45.3 8.65 5 1.7 1.3 1
cwm-selfplay_250-299 28 45.6 8.81 5 1.7 1.3 1.1
cwm-baseline-140 27.9 27.9 8.74 1 1.7 NaN NaN
cwm-selfplay-320 27.9 27.9 8.97 1 1.7 NaN NaN
cwm-baseline-230 27.9 27.9 8.72 1 1.7 NaN NaN
cwm-baseline-290 27.9 27.9 8.47 1 1.7 NaN NaN
cwm-selfplay_260-309 27.9 45.3 8.74 5 1.7 1.3 1.1
cwm-baseline-350 27.8 27.8 8.34 1 1.7 NaN NaN
cwm-selfplay-40 27.8 27.8 8.24 1 1.7 NaN NaN
cwm-baseline-330 27.8 27.8 8.37 1 1.7 NaN NaN
cwm-selfplay_270-319 27.7 44.2 8.67 5 1.7 1.3 1.1
cwm-baseline_300-349 27.7 44.6 8.41 5 1.7 1.3 1
cwm-selfplay_160-209 27.7 43.4 8.37 5 1.7 1.3 1
cwm-selfplay_100-149 27.7 44.6 8.56 5 1.7 1.3 1.1
cwm-baseline-240 27.6 27.6 8.49 1 1.7 NaN NaN
cwm-baseline-280 27.6 27.6 8.72 1 1.7 NaN NaN
cwm-selfplay-370 27.6 27.6 8.32 1 1.7 NaN NaN
cwm-selfplay-60 27.6 27.6 8.04 1 1.7 NaN NaN
cwm-selfplay-310 27.6 27.6 8.45 1 1.7 NaN NaN
cwm-selfplay_230-279 27.6 43.5 8.61 4 1.7 1.2 1.1
cwm-selfplay_220-269 27.6 42.7 8.4 4 1.7 1.3 1
cwm-baseline_180-229 27.6 44.5 8.34 5 1.7 1.3 1
cwm-selfplay_280-329 27.6 43.4 8.58 5 1.7 1.3 1.1
cwm-baseline-190 27.5 27.5 8.38 1 1.7 NaN NaN
cwm-baseline_270-319 27.5 43 8.23 5 1.7 1.3 1
cwm-selfplay_80-129 27.5 43 8.27 5 1.7 1.3 1
cwm-baseline_170-219 27.5 43.6 8.22 5 1.7 1.3 1
cwm-baseline_290-339 27.4 43.9 8.2 5 1.7 1.3 1
cwm-selfplay_180-229 27.4 43.1 8.19 5 1.6 1.3 1
cwm-baseline_280-329 27.4 43.4 8.27 5 1.6 1.3 1
cwm-selfplay_90-139 27.4 43.9 8.34 5 1.6 1.3 1
cwm-baseline_220-269 27.4 43.8 8.29 5 1.6 1.3 1
cwm-selfplay-390 27.4 27.4 8.8 1 1.6 NaN NaN
cwm-selfplay-170 27.4 27.4 8.23 1 1.6 NaN NaN
cwm-baseline_230-279 27.3 43.1 8.18 5 1.6 1.3 1
cwm-baseline_250-299 27.3 43 8.18 5 1.6 1.3 1
cwm-baseline_130-179 27.3 43.5 8.15 5 1.6 1.3 1
cwm-baseline_240-289 27.3 42.8 8.18 5 1.6 1.3 1
cwm-selfplay-190 27.2 27.2 8.26 1 1.6 NaN NaN
cwm-baseline-310 27.2 27.2 8.16 1 1.6 NaN NaN
cwm-selfplay-90 27.2 27.2 7.97 1 1.6 NaN NaN
cwm-baseline-250 27.2 27.2 8.28 1 1.6 NaN NaN
cwm-selfplay_40-89 27.2 42.8 7.88 5 1.6 1.3 1
cwm-selfplay_170-219 27.2 42.4 8.11 5 1.6 1.3 1
cwm-selfplay_50-99 27.1 42.7 7.83 5 1.6 1.3 1
cwm-baseline_260-309 27.1 42.4 7.98 5 1.6 1.3 1
cwm-selfplay_60-109 27.1 42.4 7.91 5 1.6 1.3 1
cwm-selfplay_190-239 27.1 43.1 8.08 5 1.6 1.3 1
cwm-baseline_120-169 27 43.4 8.03 5 1.6 1.3 1
cwm-selfplay_200-249 27 41.7 8.04 4 1.6 1.3 1
cwm-selfplay_210-259 27 41.6 8.05 4 1.6 1.3 1
cwm-selfplay_290-339 26.9 42.8 8.19 5 1.6 1.3 1
cwm-baseline-160 26.9 26.9 8.38 1 1.6 NaN NaN
cwm-baseline_160-209 26.9 43.4 7.95 5 1.6 1.3 1
cwm-selfplay-360 26.9 26.9 7.95 1 1.6 NaN NaN
cwm-selfplay_300-349 26.9 44 8.37 5 1.6 1.2 1.1
cwm-selfplay-200 26.9 26.9 8.36 1 1.6 NaN NaN
cwm-baseline-120 26.9 26.9 8.02 1 1.6 NaN NaN
cwm-selfplay_360-409 26.9 41.9 8.09 4 1.6 1.3 1
cwm-selfplay_310-359 26.9 44.7 8.26 5 1.6 1.2 1.1
cwm-baseline_140-189 26.8 43 7.96 5 1.6 1.3 1
cwm-selfplay-250 26.8 26.8 8.38 1 1.6 NaN NaN
cwm-selfplay-50 26.8 26.8 7.76 1 1.6 NaN NaN
cwm-selfplay-100 26.8 26.8 8.2 1 1.6 NaN NaN
cwm-baseline_100-149 26.8 43 7.83 5 1.6 1.3 1
cwm-selfplay_320-369 26.8 44.7 8.16 5 1.6 1.2 1.1
cwm-baseline_150-199 26.8 43.5 7.89 5 1.6 1.3 1
cwm-selfplay_70-119 26.8 42.1 7.8 5 1.6 1.3 1
cwm-selfplay_350-399 26.7 43.6 7.96 5 1.6 1.3 1
cwm-selfplay_330-379 26.7 43.9 8.03 5 1.6 1.3 1.1
cwm-baseline_110-159 26.7 42.5 7.77 5 1.6 1.3 1
cwm-selfplay_30-79 26.6 42.8 7.75 5 1.6 1.3 1
cwm-selfplay-230 26.5 26.5 7.98 1 1.6 NaN NaN
cwm-selfplay_340-389 26.5 43.8 7.93 5 1.6 1.2 1.1
cwm-selfplay-330 26.5 26.5 7.96 1 1.6 NaN NaN
cwm-baseline_90-139 26.5 42.8 7.71 5 1.6 1.3 1
cwm-baseline-90 26.4 26.4 8.12 1 1.6 NaN NaN
cwm-selfplay-340 26.4 26.4 8.66 1 1.6 NaN NaN
cwm-selfplay-290 26.4 26.4 7.78 1 1.6 NaN NaN
cwm-selfplay-300 26.3 26.3 8.03 1 1.6 NaN NaN
cwm-baseline-200 26.3 26.3 7.4 1 1.6 NaN NaN
cwm-selfplay_20-69 26.2 42.5 7.61 5 1.6 1.2 1
cwm-baseline-300 26 26 7.31 1 1.6 NaN NaN
cwm-selfplay-210 26 26 7.37 1 1.6 NaN NaN
cwm-selfplay-350 26 26 7.49 1 1.6 NaN NaN
cwm-selfplay-110 26 26 7.48 1 1.6 NaN NaN
cwm-selfplay-380 25.7 25.7 7.46 1 1.6 NaN NaN
cwm-baseline-180 25.7 25.7 7.17 1 1.6 NaN NaN
cwm-baseline-100 25.7 25.7 7.39 1 1.6 NaN NaN
cwm-baseline_80-129 25.6 42.1 7.4 5 1.6 1.2 1
cwm-selfplay-70 25.4 25.4 7.07 1 1.6 NaN NaN
cwm-baseline-150 25.3 25.3 7.09 1 1.6 NaN NaN
cwm-baseline-260 25.3 25.3 6.94 1 1.6 NaN NaN
cwm-baseline-40 25.3 25.3 7.46 1 1.6 NaN NaN
cwm-baseline_70-119 25.3 41.5 7.19 5 1.6 1.2 1
cwm-baseline-110 25.2 25.2 7.09 1 1.6 NaN NaN
cwm-selfplay-30 25.2 25.2 7.85 1 1.6 NaN NaN
cwm-baseline-70 25.2 25.2 6.98 1 1.6 NaN NaN
cwm-selfplay_10-59 25 41.9 7.19 5 1.6 1.2 1.1
cwm-baseline_60-109 24.9 41.3 6.98 5 1.6 1.2 1
cwm-baseline_50-99 24.4 40.2 6.7 5 1.6 1.2 1
cwm-baseline_40-89 24.1 39.9 6.57 5 1.6 1.2 1
cwm-baseline_30-79 23.9 39.5 6.45 5 1.6 1.2 1
cwm-baseline-80 23.9 23.9 6.58 1 1.6 NaN NaN
cwm-selfplay-20 23.4 23.4 6.39 1 1.6 NaN NaN
cwm-baseline_20-69 23.3 39.4 6.22 5 1.6 1.2 1
cwm-baseline-60 23.3 23.3 6.03 1 1.6 NaN NaN
cwm-baseline-50 23 23 6 1 1.6 NaN NaN
cwm-baseline-30 23 23 5.96 1 1.6 NaN NaN
cwm-baseline_10-59 22.8 39 6.1 5 1.6 1.2 1
cwm-baseline-20 22 22 5.83 1 1.5 NaN NaN
cwm-selfplay-10 22 22 5.95 1 1.5 NaN NaN
cwm-baseline-10 20.9 20.9 5.46 1 1.5 NaN NaN