######## WORLD NULL ######## === synthA === variant box textbook ideal(fit) IBM-model(fit) HARDWARE P [95%] S chi2/dof copy gap kept/circuit V11 00 +1 +1.000 +0.950 +0.926 [+0.807, +1.032] +0.010 0.30 [-0.056, +0.372] 1225 V11 01 +1 +1.000 +0.950 +0.765 [+0.649, +0.875] +0.081 1.05 [-0.269, +0.170] 1296 V11 10 +0 +0.000 +0.000 +0.080 [-0.057, +0.208] -0.021 1.21 [-0.236, +0.295] 1252 V11 11 -1 -0.854 -0.811 -0.808 [-0.914, -0.693] -0.023 3.45 [-0.078, +0.356] 1340 V10 00 +1 +1.000 +0.950 +0.976 [+0.867, +1.084] -0.014 1.02 [-0.297, +0.119] 1274 V10 01 +1 +1.000 +0.950 +0.980 [+0.876, +1.089] -0.012 0.35 [-0.390, +0.037] 1309 V10 10 -1 -0.854 -0.811 -0.883 [-0.991, -0.782] +0.017 2.14 [-0.286, +0.129] 1336 V10 11 +0 +0.000 +0.000 +0.007 [-0.125, +0.133] -0.007 0.96 [-0.114, +0.395] 1296 VA 00 -1 -0.854 -0.811 -0.871 [-0.974, -0.766] +0.005 3.84 [-0.130, +0.298] 1377 VA 01 +1 +1.000 +0.950 +0.976 [+0.869, +1.083] -0.020 1.33 [-0.078, +0.340] 1272 VA 10 +0 +0.000 +0.000 -0.009 [-0.131, +0.119] -0.006 1.22 [-0.333, +0.176] 1262 VA 11 +1 +1.000 +0.950 +0.954 [+0.844, +1.057] -0.021 2.16 [-0.246, +0.182] 1293 R1 negative box, hardware P minus IBM-model P (decision reading per variant): V11 box 11: +0.003 [-0.103, +0.118] -> CONSISTENT V10 box 10: -0.072 [-0.180, +0.030] -> CONSISTENT VA box 00: -0.060 [-0.163, +0.045] -> CONSISTENT R2 sign or position: G = (hardware/IBM-model on the negative box) minus (mean of the same ratio on the two positive boxes): V11 (negative at 11): +0.106 [-0.052, +0.263] -> CONSISTENT V10 (negative at 10): +0.060 [-0.091, +0.216] -> CONSISTENT VA (negative at 00): +0.058 [-0.088, +0.209] -> CONSISTENT R3 V11 unused box 10: S offset -0.021, P +0.080 (IBM-model P +0.000) R3 V10 unused box 11: S offset -0.007, P +0.007 (IBM-model P +0.000) R3 VA unused box 10: S offset -0.006, P -0.009 (IBM-model P +0.000) === synthB === variant box textbook ideal(fit) IBM-model(fit) HARDWARE P [95%] S chi2/dof copy gap kept/circuit V11 00 +1 +1.000 +0.950 +0.897 [+0.793, +1.012] +0.009 0.52 [-0.239, +0.205] 1293 V11 01 +1 +1.000 +0.950 +1.062 [+0.956, +1.165] -0.042 1.53 [-0.298, +0.129] 1248 V11 10 +0 +0.000 +0.000 -0.056 [-0.186, +0.067] +0.034 1.23 [-0.273, +0.262] 1285 V11 11 -1 -0.854 -0.811 -0.950 [-1.053, -0.848] +0.050 0.55 [-0.257, +0.169] 1353 V10 00 +1 +1.000 +0.950 +0.885 [+0.779, +1.000] +0.027 0.59 [-0.223, +0.211] 1287 V10 01 +1 +1.000 +0.950 +0.828 [+0.719, +0.935] +0.051 1.56 [-0.309, +0.143] 1289 V10 10 -1 -0.854 -0.811 -0.705 [-0.815, -0.600] -0.077 0.23 [-0.212, +0.225] 1391 V10 11 +0 +0.000 +0.000 +0.029 [-0.098, +0.163] -0.034 2.86 [+0.018, +0.516] 1286 VA 00 -1 -0.854 -0.811 -0.764 [-0.873, -0.663] -0.056 0.75 [+0.010, +0.430] 1348 VA 01 +1 +1.000 +0.950 +0.976 [+0.869, +1.086] -0.016 0.12 [-0.043, +0.389] 1328 VA 10 +0 +0.000 +0.000 +0.159 [+0.038, +0.289] -0.078 0.75 [-0.260, +0.273] 1281 VA 11 +1 +1.000 +0.950 +0.993 [+0.884, +1.096] -0.034 1.70 [-0.101, +0.348] 1256 R1 negative box, hardware P minus IBM-model P (decision reading per variant): V11 box 11: -0.139 [-0.241, -0.037] -> BELOW V10 box 10: +0.106 [-0.004, +0.211] -> CONSISTENT VA box 00: +0.047 [-0.062, +0.148] -> CANNOT_TELL (copies disagree) R2 sign or position: G = (hardware/IBM-model on the negative box) minus (mean of the same ratio on the two positive boxes): V11 (negative at 11): +0.140 [-0.014, +0.284] -> CONSISTENT V10 (negative at 10): -0.032 [-0.182, +0.124] -> CONSISTENT VA (negative at 00): -0.095 [-0.248, +0.056] -> CANNOT_TELL (copies disagree) R3 V11 unused box 10: S offset +0.034, P -0.056 (IBM-model P +0.000) R3 V10 unused box 11: S offset -0.034, P +0.029 (IBM-model P +0.000) R3 VA unused box 10: S offset -0.078, P +0.159 (IBM-model P +0.000) === synthC === variant box textbook ideal(fit) IBM-model(fit) HARDWARE P [95%] S chi2/dof copy gap kept/circuit V11 00 +1 +1.000 +0.950 +1.026 [+0.922, +1.128] -0.043 1.67 [+0.010, +0.429] 1328 V11 01 +1 +1.000 +0.950 +0.940 [+0.830, +1.053] -0.006 0.99 [-0.035, +0.408] 1276 V11 10 +0 +0.000 +0.000 +0.058 [-0.069, +0.181] -0.008 1.67 [-0.276, +0.246] 1289 V11 11 -1 -0.854 -0.811 -0.781 [-0.887, -0.675] -0.047 0.40 [-0.153, +0.278] 1341 V10 00 +1 +1.000 +0.950 +1.006 [+0.895, +1.115] -0.033 0.70 [-0.296, +0.136] 1288 V10 01 +1 +1.000 +0.950 +0.937 [+0.833, +1.047] -0.009 1.40 [-0.310, +0.137] 1309 V10 10 -1 -0.854 -0.811 -0.821 [-0.921, -0.719] -0.027 0.55 [-0.295, +0.114] 1357 V10 11 +0 +0.000 +0.000 +0.019 [-0.104, +0.151] -0.010 1.20 [-0.407, +0.118] 1281 VA 00 -1 -0.854 -0.811 -0.780 [-0.881, -0.678] -0.052 1.01 [-0.194, +0.221] 1340 VA 01 +1 +1.000 +0.950 +0.927 [+0.823, +1.035] +0.016 0.10 [+0.015, +0.462] 1294 VA 10 +0 +0.000 +0.000 +0.033 [-0.093, +0.153] -0.025 1.06 [-0.193, +0.287] 1309 VA 11 +1 +1.000 +0.950 +1.032 [+0.926, +1.144] -0.037 0.35 [-0.336, +0.112] 1263 R1 negative box, hardware P minus IBM-model P (decision reading per variant): V11 box 11: +0.030 [-0.076, +0.136] -> CONSISTENT V10 box 10: -0.010 [-0.110, +0.092] -> CONSISTENT VA box 00: +0.032 [-0.070, +0.133] -> CONSISTENT R2 sign or position: G = (hardware/IBM-model on the negative box) minus (mean of the same ratio on the two positive boxes): V11 (negative at 11): -0.071 [-0.224, +0.084] -> CANNOT_TELL (copies disagree) V10 (negative at 10): -0.010 [-0.162, +0.143] -> CONSISTENT VA (negative at 00): -0.070 [-0.223, +0.081] -> CANNOT_TELL (copies disagree) R3 V11 unused box 10: S offset -0.008, P +0.058 (IBM-model P +0.000) R3 V10 unused box 11: S offset -0.010, P +0.019 (IBM-model P +0.000) R3 VA unused box 10: S offset -0.025, P +0.033 (IBM-model P +0.000) COMBINED across chips (R2, per variant; sigma = 95% width / 3.92): V11: +0.059 [-0.029, +0.148] chi2/dof 2.11 -> not combined (chip-to-chip chi2/dof >= 2) V10: +0.006 [-0.083, +0.094] chi2/dof 0.38 -> CONSISTENT VA: -0.034 [-0.121, +0.053] chi2/dof 1.16 -> CONSISTENT ######## WORLD SIGN ######## === synthA === variant box textbook ideal(fit) IBM-model(fit) HARDWARE P [95%] S chi2/dof copy gap kept/circuit V11 00 +1 +1.000 +0.950 +0.902 [+0.788, +1.011] -0.039 1.05 [-0.019, +0.421] 1274 V11 01 +1 +1.000 +0.950 +0.764 [+0.651, +0.878] +0.040 0.46 [-0.187, +0.279] 1272 V11 10 +0 +0.000 +0.000 +0.070 [-0.059, +0.198] -0.042 0.23 [-0.258, +0.271] 1294 V11 11 -1 -0.854 -0.811 -0.475 [-0.587, -0.360] -0.001 1.30 [-0.275, +0.199] 1373 V10 00 +1 +1.000 +0.950 +0.888 [+0.773, +0.998] -0.009 1.85 [-0.329, +0.119] 1310 V10 01 +1 +1.000 +0.950 +0.935 [+0.818, +1.049] -0.037 0.35 [-0.368, +0.078] 1287 V10 10 -1 -0.854 -0.811 -0.413 [-0.528, -0.304] -0.042 0.90 [-0.185, +0.263] 1357 V10 11 +0 +0.000 +0.000 +0.084 [-0.048, +0.213] -0.040 1.20 [-0.076, +0.447] 1247 VA 00 -1 -0.854 -0.811 -0.414 [-0.526, -0.299] -0.037 0.22 [-0.219, +0.240] 1377 VA 01 +1 +1.000 +0.950 +0.872 [+0.761, +0.990] -0.005 3.25 [-0.399, +0.035] 1309 VA 10 +0 +0.000 +0.000 +0.052 [-0.083, +0.182] -0.023 0.81 [-0.473, +0.044] 1283 VA 11 +1 +1.000 +0.950 +0.898 [+0.787, +1.012] +0.005 2.07 [+0.024, +0.465] 1282 R1 negative box, hardware P minus IBM-model P (decision reading per variant): V11 box 11: +0.336 [+0.224, +0.451] -> ABOVE V10 box 10: +0.398 [+0.283, +0.507] -> ABOVE VA box 00: +0.397 [+0.285, +0.512] -> ABOVE R2 sign or position: G = (hardware/IBM-model on the negative box) minus (mean of the same ratio on the two positive boxes): V11 (negative at 11): -0.292 [-0.461, -0.131] -> BELOW V10 (negative at 10): -0.450 [-0.613, -0.279] -> BELOW VA (negative at 00): -0.421 [-0.584, -0.259] -> CANNOT_TELL (copies disagree) R3 V11 unused box 10: S offset -0.042, P +0.070 (IBM-model P +0.000) R3 V10 unused box 11: S offset -0.040, P +0.084 (IBM-model P +0.000) R3 VA unused box 10: S offset -0.023, P +0.052 (IBM-model P +0.000) === synthB === variant box textbook ideal(fit) IBM-model(fit) HARDWARE P [95%] S chi2/dof copy gap kept/circuit V11 00 +1 +1.000 +0.950 +0.833 [+0.720, +0.949] -0.009 0.90 [-0.213, +0.250] 1290 V11 01 +1 +1.000 +0.950 +0.836 [+0.726, +0.947] +0.015 1.52 [+0.034, +0.475] 1291 V11 10 +0 +0.000 +0.000 -0.112 [-0.240, +0.024] +0.061 0.38 [-0.188, +0.319] 1312 V11 11 -1 -0.854 -0.811 -0.365 [-0.479, -0.248] -0.069 0.42 [-0.167, +0.288] 1369 V10 00 +1 +1.000 +0.950 +0.806 [+0.693, +0.917] +0.024 3.43 [-0.284, +0.182] 1299 V10 01 +1 +1.000 +0.950 +0.891 [+0.780, +1.005] -0.024 0.30 [-0.167, +0.275] 1324 V10 10 -1 -0.854 -0.811 -0.446 [-0.562, -0.334] -0.032 1.28 [-0.036, +0.431] 1365 V10 11 +0 +0.000 +0.000 -0.043 [-0.174, +0.082] +0.013 1.47 [-0.224, +0.304] 1327 VA 00 -1 -0.854 -0.811 -0.424 [-0.538, -0.316] -0.027 1.79 [-0.259, +0.171] 1393 VA 01 +1 +1.000 +0.950 +0.955 [+0.843, +1.068] -0.060 0.82 [-0.232, +0.217] 1271 VA 10 +0 +0.000 +0.000 -0.009 [-0.132, +0.120] -0.001 0.59 [-0.142, +0.366] 1264 VA 11 +1 +1.000 +0.950 +0.847 [+0.742, +0.963] +0.007 0.98 [-0.334, +0.109] 1266 R1 negative box, hardware P minus IBM-model P (decision reading per variant): V11 box 11: +0.446 [+0.332, +0.563] -> ABOVE V10 box 10: +0.366 [+0.249, +0.477] -> ABOVE VA box 00: +0.387 [+0.273, +0.495] -> ABOVE R2 sign or position: G = (hardware/IBM-model on the negative box) minus (mean of the same ratio on the two positive boxes): V11 (negative at 11): -0.429 [-0.599, -0.263] -> CANNOT_TELL (copies disagree) V10 (negative at 10): -0.344 [-0.506, -0.184] -> BELOW VA (negative at 00): -0.426 [-0.593, -0.261] -> BELOW R3 V11 unused box 10: S offset +0.061, P -0.112 (IBM-model P +0.000) R3 V10 unused box 11: S offset +0.013, P -0.043 (IBM-model P +0.000) R3 VA unused box 10: S offset -0.001, P -0.009 (IBM-model P +0.000) === synthC === variant box textbook ideal(fit) IBM-model(fit) HARDWARE P [95%] S chi2/dof copy gap kept/circuit V11 00 +1 +1.000 +0.950 +0.850 [+0.736, +0.963] +0.001 1.15 [-0.297, +0.179] 1286 V11 01 +1 +1.000 +0.950 +0.824 [+0.715, +0.939] +0.010 0.24 [-0.169, +0.293] 1288 V11 10 +0 +0.000 +0.000 +0.000 [-0.129, +0.124] -0.002 0.15 [-0.573, -0.071] 1302 V11 11 -1 -0.854 -0.811 -0.437 [-0.553, -0.318] -0.049 1.07 [-0.366, +0.094] 1346 V10 00 +1 +1.000 +0.950 +0.857 [+0.744, +0.976] -0.010 1.15 [-0.303, +0.156] 1279 V10 01 +1 +1.000 +0.950 +0.847 [+0.730, +0.966] +0.007 0.49 [-0.417, +0.037] 1286 V10 10 -1 -0.854 -0.811 -0.410 [-0.522, -0.289] -0.057 1.74 [-0.229, +0.224] 1295 V10 11 +0 +0.000 +0.000 +0.047 [-0.082, +0.179] -0.035 0.23 [+0.155, +0.677] 1306 VA 00 -1 -0.854 -0.811 -0.494 [-0.608, -0.383] -0.011 0.47 [-0.053, +0.395] 1384 VA 01 +1 +1.000 +0.950 +0.808 [+0.695, +0.920] +0.023 0.55 [-0.412, +0.058] 1238 VA 10 +0 +0.000 +0.000 +0.051 [-0.079, +0.181] -0.020 0.41 [-0.343, +0.162] 1265 VA 11 +1 +1.000 +0.950 +0.862 [+0.752, +0.968] -0.013 0.27 [-0.308, +0.153] 1294 R1 negative box, hardware P minus IBM-model P (decision reading per variant): V11 box 11: +0.374 [+0.258, +0.493] -> ABOVE V10 box 10: +0.401 [+0.289, +0.522] -> ABOVE VA box 00: +0.317 [+0.203, +0.428] -> ABOVE R2 sign or position: G = (hardware/IBM-model on the negative box) minus (mean of the same ratio on the two positive boxes): V11 (negative at 11): -0.342 [-0.514, -0.182] -> BELOW V10 (negative at 10): -0.392 [-0.557, -0.233] -> BELOW VA (negative at 00): -0.270 [-0.431, -0.109] -> BELOW R3 V11 unused box 10: S offset -0.002, P +0.000 (IBM-model P +0.000) R3 V10 unused box 11: S offset -0.035, P +0.047 (IBM-model P +0.000) R3 VA unused box 10: S offset -0.020, P +0.051 (IBM-model P +0.000) COMBINED across chips (R2, per variant; sigma = 95% width / 3.92): V11: -0.353 [-0.449, -0.257] chi2/dof 0.67 -> BELOW V10: -0.394 [-0.488, -0.300] chi2/dof 0.40 -> BELOW VA: -0.371 [-0.465, -0.277] chi2/dof 1.15 -> BELOW ######## WORLD CORNER ######## === synthA === variant box textbook ideal(fit) IBM-model(fit) HARDWARE P [95%] S chi2/dof copy gap kept/circuit V11 00 +1 +1.000 +0.950 +0.799 [+0.685, +0.910] +0.030 0.44 [-0.229, +0.222] 1314 V11 01 +1 +1.000 +0.950 +0.852 [+0.734, +0.965] -0.011 0.56 [-0.070, +0.393] 1302 V11 10 +0 +0.000 +0.000 +0.022 [-0.100, +0.144] -0.008 0.21 [-0.536, -0.002] 1280 V11 11 -1 -0.854 -0.811 -0.335 [-0.455, -0.220] -0.060 0.66 [-0.046, +0.401] 1334 V10 00 +1 +1.000 +0.950 +0.832 [+0.720, +0.944] +0.012 0.30 [-0.281, +0.184] 1267 V10 01 +1 +1.000 +0.950 +0.834 [+0.721, +0.949] +0.009 0.43 [-0.125, +0.320] 1266 V10 10 -1 -0.854 -0.811 -0.715 [-0.825, -0.604] -0.033 0.29 [-0.134, +0.301] 1350 V10 11 +0 +0.000 +0.000 -0.094 [-0.230, +0.040] +0.040 0.04 [-0.269, +0.244] 1274 VA 00 -1 -0.854 -0.811 -0.719 [-0.828, -0.610] -0.015 0.75 [-0.393, +0.037] 1285 VA 01 +1 +1.000 +0.950 +0.863 [+0.754, +0.974] -0.003 0.02 [-0.258, +0.184] 1276 VA 10 +0 +0.000 +0.000 -0.038 [-0.164, +0.085] +0.016 0.25 [-0.342, +0.168] 1300 VA 11 +1 +1.000 +0.950 +0.456 [+0.334, +0.578] +0.037 0.71 [-0.410, +0.076] 1334 R1 negative box, hardware P minus IBM-model P (decision reading per variant): V11 box 11: +0.477 [+0.356, +0.591] -> ABOVE V10 box 10: +0.096 [-0.014, +0.207] -> CONSISTENT VA box 00: +0.093 [-0.017, +0.201] -> CONSISTENT R2 sign or position: G = (hardware/IBM-model on the negative box) minus (mean of the same ratio on the two positive boxes): V11 (negative at 11): -0.456 [-0.620, -0.292] -> BELOW V10 (negative at 10): +0.004 [-0.149, +0.166] -> CONSISTENT VA (negative at 00): +0.192 [+0.027, +0.354] -> ABOVE R3 V11 unused box 10: S offset -0.008, P +0.022 (IBM-model P +0.000) R3 V10 unused box 11: S offset +0.040, P -0.094 (IBM-model P +0.000) R3 VA unused box 10: S offset +0.016, P -0.038 (IBM-model P +0.000) === synthB === variant box textbook ideal(fit) IBM-model(fit) HARDWARE P [95%] S chi2/dof copy gap kept/circuit V11 00 +1 +1.000 +0.950 +0.920 [+0.807, +1.033] -0.022 1.15 [-0.165, +0.274] 1261 V11 01 +1 +1.000 +0.950 +0.867 [+0.752, +0.976] +0.009 0.83 [+0.007, +0.439] 1277 V11 10 +0 +0.000 +0.000 -0.074 [-0.210, +0.057] +0.023 0.68 [-0.215, +0.294] 1268 V11 11 -1 -0.854 -0.811 -0.489 [-0.605, -0.372] -0.004 0.25 [-0.131, +0.323] 1402 V10 00 +1 +1.000 +0.950 +0.903 [+0.785, +1.015] -0.056 0.30 [-0.437, +0.036] 1299 V10 01 +1 +1.000 +0.950 +0.918 [+0.808, +1.029] -0.021 1.17 [-0.161, +0.288] 1313 V10 10 -1 -0.854 -0.811 -0.731 [-0.837, -0.618] -0.040 1.36 [-0.050, +0.379] 1313 V10 11 +0 +0.000 +0.000 +0.003 [-0.121, +0.128] -0.001 0.56 [-0.392, +0.125] 1287 VA 00 -1 -0.854 -0.811 -0.724 [-0.833, -0.614] -0.008 2.59 [-0.181, +0.261] 1359 VA 01 +1 +1.000 +0.950 +0.817 [+0.700, +0.927] +0.023 1.16 [-0.226, +0.217] 1261 VA 10 +0 +0.000 +0.000 -0.108 [-0.242, +0.015] +0.048 0.68 [-0.511, -0.006] 1291 VA 11 +1 +1.000 +0.950 +0.468 [+0.342, +0.590] +0.042 0.29 [-0.372, +0.119] 1303 R1 negative box, hardware P minus IBM-model P (decision reading per variant): V11 box 11: +0.322 [+0.206, +0.439] -> ABOVE V10 box 10: +0.080 [-0.026, +0.193] -> CONSISTENT VA box 00: +0.087 [-0.022, +0.197] -> CONSISTENT R2 sign or position: G = (hardware/IBM-model on the negative box) minus (mean of the same ratio on the two positive boxes): V11 (negative at 11): -0.338 [-0.507, -0.173] -> CANNOT_TELL (copies disagree) V10 (negative at 10): -0.057 [-0.214, +0.102] -> CONSISTENT VA (negative at 00): +0.216 [+0.050, +0.385] -> ABOVE R3 V11 unused box 10: S offset +0.023, P -0.074 (IBM-model P +0.000) R3 V10 unused box 11: S offset -0.001, P +0.003 (IBM-model P +0.000) R3 VA unused box 10: S offset +0.048, P -0.108 (IBM-model P +0.000) === synthC === variant box textbook ideal(fit) IBM-model(fit) HARDWARE P [95%] S chi2/dof copy gap kept/circuit V11 00 +1 +1.000 +0.950 +0.859 [+0.749, +0.976] -0.005 1.27 [-0.258, +0.185] 1279 V11 01 +1 +1.000 +0.950 +0.827 [+0.720, +0.937] +0.013 0.79 [-0.149, +0.293] 1219 V11 10 +0 +0.000 +0.000 +0.024 [-0.097, +0.151] -0.011 0.72 [-0.268, +0.233] 1273 V11 11 -1 -0.854 -0.811 -0.518 [-0.632, -0.403] +0.014 2.37 [-0.322, +0.135] 1384 V10 00 +1 +1.000 +0.950 +0.865 [+0.752, +0.983] -0.031 0.53 [-0.123, +0.336] 1310 V10 01 +1 +1.000 +0.950 +0.785 [+0.665, +0.909] +0.027 0.25 [-0.353, +0.117] 1246 V10 10 -1 -0.854 -0.811 -0.828 [-0.942, -0.715] +0.024 0.54 [-0.204, +0.229] 1347 V10 11 +0 +0.000 +0.000 -0.036 [-0.163, +0.090] +0.014 1.85 [-0.316, +0.186] 1292 VA 00 -1 -0.854 -0.811 -0.724 [-0.825, -0.618] -0.059 0.73 [-0.238, +0.174] 1337 VA 01 +1 +1.000 +0.950 +0.828 [+0.708, +0.942] -0.004 0.69 [-0.210, +0.250] 1249 VA 10 +0 +0.000 +0.000 -0.097 [-0.225, +0.034] +0.045 1.06 [-0.146, +0.376] 1267 VA 11 +1 +1.000 +0.950 +0.569 [+0.447, +0.685] -0.026 1.41 [-0.168, +0.322] 1310 R1 negative box, hardware P minus IBM-model P (decision reading per variant): V11 box 11: +0.293 [+0.180, +0.408] -> ABOVE V10 box 10: -0.017 [-0.131, +0.096] -> CONSISTENT VA box 00: +0.087 [-0.014, +0.193] -> CONSISTENT R2 sign or position: G = (hardware/IBM-model on the negative box) minus (mean of the same ratio on the two positive boxes): V11 (negative at 11): -0.249 [-0.404, -0.085] -> BELOW V10 (negative at 10): +0.153 [-0.018, +0.313] -> CONSISTENT VA (negative at 00): +0.158 [-0.003, +0.313] -> CONSISTENT R3 V11 unused box 10: S offset -0.011, P +0.024 (IBM-model P +0.000) R3 V10 unused box 11: S offset +0.014, P -0.036 (IBM-model P +0.000) R3 VA unused box 10: S offset +0.045, P -0.097 (IBM-model P +0.000) COMBINED across chips (R2, per variant; sigma = 95% width / 3.92): V11: -0.346 [-0.440, -0.251] chi2/dof 1.58 -> BELOW V10: +0.029 [-0.063, +0.122] chi2/dof 1.69 -> CONSISTENT VA: +0.188 [+0.093, +0.282] chi2/dof 0.13 -> ABOVE