############################## SCENARIO a seed 20261220 ==================================================================================================== ibm_marrakesh job SYNTH-a-20261220 CAL 0X flip-rule failure per measured qubit: median 1.95%, max 98.64%; flip-unreadable (> 20%): [82, 100]; wired spectators 86, used for L1/L2/L3 83 KNOWN-GOOD 0X: flip success, median over measured qubits 98.05% (flip-readable 98.08%) -> PASS F1 N=0 idle copy 0 fail 2.9541% (inside 0 2.7495%, outside both 0.2045%) | with CAL1 cluster 2.9545% F1 N=0 idle copy 1 fail 2.9568% (inside 0 2.7387%, outside both 0.2180%) | with CAL1 cluster 2.9586% F1 N=0 idle pooled fail 2.9554% (inside 0 2.7441%, outside both 0.2113%) | with CAL1 cluster 2.9565% F1 N=4 idle copy 0 fail 2.9955% (inside 0 2.7694%, outside both 0.2261%) | with CAL1 cluster 2.9892% F1 N=4 idle copy 1 fail 3.0050% (inside 0 2.7941%, outside both 0.2108%) | with CAL1 cluster 3.0077% F1 N=4 idle pooled fail 3.0002% (inside 0 2.7818%, outside both 0.2185%) | with CAL1 cluster 2.9984% F1 N=4 readout copy 0 fail 3.4221% (inside 0 3.2023%, outside both 0.2198%) | with CAL1 cluster 3.4266% | among shots verified 0 on every mid read 2.9483% F1 N=4 readout copy 1 fail 3.4694% (inside 0 3.2653%, outside both 0.2041%) | with CAL1 cluster 3.4703% | among shots verified 0 on every mid read 2.9990% F1 N=4 readout pooled fail 3.4457% (inside 0 3.2338%, outside both 0.2119%) | with CAL1 cluster 3.4484% | among shots verified 0 on every mid read 2.9737% F1 N=8 idle copy 0 fail 3.0149% (inside 0 2.7896%, outside both 0.2252%) | with CAL1 cluster 3.0126% F1 N=8 idle copy 1 fail 2.9045% (inside 0 2.6856%, outside both 0.2189%) | with CAL1 cluster 2.9050% F1 N=8 idle pooled fail 2.9597% (inside 0 2.7376%, outside both 0.2221%) | with CAL1 cluster 2.9588% F1 N=8 readout copy 0 fail 3.4959% (inside 0 3.2883%, outside both 0.2077%) | with CAL1 cluster 3.4982% | among shots verified 0 on every mid read 2.9928% F1 N=8 readout copy 1 fail 3.3820% (inside 0 3.1743%, outside both 0.2077%) | with CAL1 cluster 3.3851% | among shots verified 0 on every mid read 2.9142% F1 N=8 readout pooled fail 3.4390% (inside 0 3.2313%, outside both 0.2077%) | with CAL1 cluster 3.4417% | among shots verified 0 on every mid read 2.9535% KNOWN-GOOD idle N=0: flip failure minus CAL 0X floor -0.0395% [-0.1080, +0.0330] -> CONSISTENT -> PASS KNOWN-GOOD idle N=4: flip failure minus CAL 0X floor +0.0053% [-0.0758, +0.0915] -> CONSISTENT -> PASS KNOWN-GOOD idle N=8: flip failure minus CAL 0X floor -0.0352% [-0.1179, +0.0419] -> CANNOT_TELL (copies disagree) -> PASS N=4 F2 hidden-leak rate (readout, verified-0 shots) - idle copy 0 -0.0472% [-0.1393, +0.0486] n=37 -> CONSISTENT N=4 F2 hidden-leak rate (readout, verified-0 shots) - idle copy 1 -0.0059% [-0.0922, +0.0824] n=37 -> CONSISTENT N=4 F2 hidden-leak rate (readout, verified-0 shots) - idle pooled -0.0266% [-0.0918, +0.0392] n=37 -> CONSISTENT N=4 F2 FINAL -> CONSISTENT N=4 F2u all-shot flip failure, readout - idle copy 0 +0.4266% [+0.3243, +0.5275] n=37 -> ABOVE N=4 F2u all-shot flip failure, readout - idle copy 1 +0.4644% [+0.3743, +0.5514] n=37 -> ABOVE N=4 F2u all-shot flip failure, readout - idle pooled +0.4455% [+0.3818, +0.5126] n=37 -> ABOVE N=4 F2u FINAL -> ABOVE N=4 L1w loss, mid-verified-0, ALL wired (run_iq R1 repeat) copy 0 +0.0075 [+0.0022, +0.0108] n=86 -> ABOVE N=4 L1w loss, mid-verified-0, ALL wired (run_iq R1 repeat) copy 1 +0.0082 [+0.0049, +0.0122] n=86 -> ABOVE N=4 L1w loss, mid-verified-0, ALL wired (run_iq R1 repeat) pooled +0.0089 [+0.0048, +0.0124] n=86 -> ABOVE N=4 L1w FINAL -> ABOVE N=4 L1 loss, mid-verified-0 selection copy 0 +0.0079 [+0.0040, +0.0109] n=83 -> ABOVE N=4 L1 loss, mid-verified-0 selection copy 1 +0.0084 [+0.0046, +0.0132] n=83 -> ABOVE N=4 L1 loss, mid-verified-0 selection pooled +0.0098 [+0.0053, +0.0127] n=83 -> ABOVE N=4 L1 FINAL -> ABOVE N=4 L2 loss, mid-verified-0 AND flipped copy 0 +0.0082 [+0.0048, +0.0124] n=83 -> ABOVE N=4 L2 loss, mid-verified-0 AND flipped copy 1 +0.0080 [+0.0050, +0.0127] n=83 -> ABOVE N=4 L2 loss, mid-verified-0 AND flipped pooled +0.0088 [+0.0050, +0.0121] n=83 -> ABOVE N=4 L2 FINAL -> ABOVE N=4 L3 L2 - L1 copy 0 +0.0003 [-0.0003, +0.0008] n=83 -> CONSISTENT N=4 L3 L2 - L1 copy 1 -0.0002 [-0.0007, +0.0004] n=83 -> CONSISTENT N=4 L3 L2 - L1 pooled +0.0000 [-0.0004, +0.0004] n=83 -> CONSISTENT N=4 L3 FINAL -> CONSISTENT KNOWN-GOOD N=4: L1w +0.0089 vs 21 Oct clean-shot loss +0.0087 [+0.0032, +0.0160] -> PASS N=4 --- marrakesh without neighbours of overlap/miss qubits (printed beside, does not replace): spectators used 81 (L1/L2/L3), 81 (L1w); dropped neighbours of [82, 100, 119] N=4 SIDE L1w copy 0 +0.0079 [+0.0040, +0.0109] n=81 -> ABOVE N=4 SIDE L1w copy 1 +0.0084 [+0.0046, +0.0132] n=81 -> ABOVE N=4 SIDE L1w pooled +0.0098 [+0.0053, +0.0127] n=81 -> ABOVE N=4 SIDE L1w FINAL -> ABOVE N=4 SIDE L1 copy 0 +0.0079 [+0.0040, +0.0109] n=81 -> ABOVE N=4 SIDE L1 copy 1 +0.0084 [+0.0046, +0.0131] n=81 -> ABOVE N=4 SIDE L1 pooled +0.0098 [+0.0042, +0.0127] n=81 -> ABOVE N=4 SIDE L1 FINAL -> ABOVE N=4 SIDE L2 copy 0 +0.0082 [+0.0045, +0.0124] n=81 -> ABOVE N=4 SIDE L2 copy 1 +0.0080 [+0.0050, +0.0127] n=81 -> ABOVE N=4 SIDE L2 pooled +0.0089 [+0.0050, +0.0121] n=81 -> ABOVE N=4 SIDE L2 FINAL -> ABOVE N=4 SIDE L3 copy 0 +0.0003 [-0.0003, +0.0008] n=81 -> CONSISTENT N=4 SIDE L3 copy 1 +0.0000 [-0.0007, +0.0006] n=81 -> CONSISTENT N=4 SIDE L3 pooled +0.0000 [-0.0004, +0.0005] n=81 -> CONSISTENT N=4 SIDE L3 FINAL -> CONSISTENT KNOWN-GOOD SIDE N=4: L1w without those neighbours +0.0098 vs 21 Oct clean-shot loss +0.0087 [+0.0032, +0.0160] -> PASS N=4 shots kept (median over wired spectators, pooled): L1 readout 98.6%, L2 readout 96.7%, L2 idle 98.0% N=4 DECISION -> L2 ABOVE: a neighbour stuck in a level that reads as 0 is EXCLUDED as the cause (to the extent the X-flip catches it) | L3 CONSISTENT, F2 CONSISTENT N=8 F2 hidden-leak rate (readout, verified-0 shots) - idle copy 0 -0.0220% [-0.1033, +0.0597] n=37 -> CONSISTENT N=8 F2 hidden-leak rate (readout, verified-0 shots) - idle copy 1 +0.0097% [-0.0917, +0.1135] n=37 -> CONSISTENT N=8 F2 hidden-leak rate (readout, verified-0 shots) - idle pooled -0.0062% [-0.0716, +0.0629] n=37 -> CONSISTENT N=8 F2 FINAL -> CONSISTENT N=8 F2u all-shot flip failure, readout - idle copy 0 +0.4811% [+0.3986, +0.5595] n=37 -> ABOVE N=8 F2u all-shot flip failure, readout - idle copy 1 +0.4775% [+0.3757, +0.5834] n=37 -> ABOVE N=8 F2u all-shot flip failure, readout - idle pooled +0.4793% [+0.4149, +0.5536] n=37 -> ABOVE N=8 F2u FINAL -> ABOVE N=8 L1w loss, mid-verified-0, ALL wired (run_iq R1 repeat) copy 0 +0.0095 [+0.0036, +0.0114] n=86 -> ABOVE N=8 L1w loss, mid-verified-0, ALL wired (run_iq R1 repeat) copy 1 +0.0102 [+0.0058, +0.0136] n=86 -> ABOVE N=8 L1w loss, mid-verified-0, ALL wired (run_iq R1 repeat) pooled +0.0096 [+0.0059, +0.0117] n=86 -> ABOVE N=8 L1w FINAL -> ABOVE N=8 L1 loss, mid-verified-0 selection copy 0 +0.0095 [+0.0038, +0.0114] n=83 -> ABOVE N=8 L1 loss, mid-verified-0 selection copy 1 +0.0102 [+0.0053, +0.0136] n=83 -> ABOVE N=8 L1 loss, mid-verified-0 selection pooled +0.0094 [+0.0057, +0.0115] n=83 -> ABOVE N=8 L1 FINAL -> ABOVE N=8 L2 loss, mid-verified-0 AND flipped copy 0 +0.0078 [+0.0032, +0.0115] n=83 -> ABOVE N=8 L2 loss, mid-verified-0 AND flipped copy 1 +0.0101 [+0.0059, +0.0128] n=83 -> ABOVE N=8 L2 loss, mid-verified-0 AND flipped pooled +0.0097 [+0.0042, +0.0114] n=83 -> ABOVE N=8 L2 FINAL -> ABOVE N=8 L3 L2 - L1 copy 0 +0.0002 [-0.0005, +0.0006] n=83 -> CONSISTENT N=8 L3 L2 - L1 copy 1 -0.0003 [-0.0009, +0.0005] n=83 -> CONSISTENT N=8 L3 L2 - L1 pooled +0.0000 [-0.0007, +0.0004] n=83 -> CONSISTENT N=8 L3 FINAL -> CONSISTENT KNOWN-GOOD N=8: L1w +0.0096 vs 21 Oct clean-shot loss +0.0205 [+0.0112, +0.0284] -> FAIL N=8 --- marrakesh without neighbours of overlap/miss qubits (printed beside, does not replace): spectators used 81 (L1/L2/L3), 81 (L1w); dropped neighbours of [82, 100, 119] N=8 SIDE L1w copy 0 +0.0095 [+0.0036, +0.0114] n=81 -> ABOVE N=8 SIDE L1w copy 1 +0.0102 [+0.0053, +0.0136] n=81 -> ABOVE N=8 SIDE L1w pooled +0.0094 [+0.0057, +0.0113] n=81 -> ABOVE N=8 SIDE L1w FINAL -> ABOVE N=8 SIDE L1 copy 0 +0.0095 [+0.0034, +0.0114] n=81 -> ABOVE N=8 SIDE L1 copy 1 +0.0102 [+0.0055, +0.0136] n=81 -> ABOVE N=8 SIDE L1 pooled +0.0094 [+0.0057, +0.0115] n=81 -> ABOVE N=8 SIDE L1 FINAL -> ABOVE N=8 SIDE L2 copy 0 +0.0074 [+0.0029, +0.0115] n=81 -> ABOVE N=8 SIDE L2 copy 1 +0.0101 [+0.0059, +0.0128] n=81 -> ABOVE N=8 SIDE L2 pooled +0.0097 [+0.0042, +0.0114] n=81 -> ABOVE N=8 SIDE L2 FINAL -> ABOVE N=8 SIDE L3 copy 0 +0.0002 [-0.0005, +0.0006] n=81 -> CONSISTENT N=8 SIDE L3 copy 1 -0.0005 [-0.0012, +0.0006] n=81 -> CONSISTENT N=8 SIDE L3 pooled +0.0000 [-0.0007, +0.0004] n=81 -> CONSISTENT N=8 SIDE L3 FINAL -> CONSISTENT KNOWN-GOOD SIDE N=8: L1w without those neighbours +0.0094 vs 21 Oct clean-shot loss +0.0205 [+0.0112, +0.0284] -> FAIL N=8 shots kept (median over wired spectators, pooled): L1 readout 97.8%, L2 readout 95.8%, L2 idle 98.1% N=8 DECISION -> L2 ABOVE: a neighbour stuck in a level that reads as 0 is EXCLUDED as the cause (to the extent the X-flip catches it) | L3 CONSISTENT, F2 CONSISTENT ############################## SCENARIO b seed 20261221 ==================================================================================================== ibm_marrakesh job SYNTH-b-20261221 CAL 0X flip-rule failure per measured qubit: median 1.91%, max 98.76%; flip-unreadable (> 20%): [82, 100]; wired spectators 86, used for L1/L2/L3 83 KNOWN-GOOD 0X: flip success, median over measured qubits 98.09% (flip-readable 98.12%) -> PASS F1 N=0 idle copy 0 fail 2.8878% (inside 0 2.6955%, outside both 0.1923%) | with CAL1 cluster 2.8887% F1 N=0 idle copy 1 fail 2.8432% (inside 0 2.6158%, outside both 0.2275%) | with CAL1 cluster 2.8378% F1 N=0 idle pooled fail 2.8655% (inside 0 2.6556%, outside both 0.2099%) | with CAL1 cluster 2.8633% F1 N=4 idle copy 0 fail 2.8086% (inside 0 2.5883%, outside both 0.2203%) | with CAL1 cluster 2.8063% F1 N=4 idle copy 1 fail 2.9045% (inside 0 2.6820%, outside both 0.2225%) | with CAL1 cluster 2.8914% F1 N=4 idle pooled fail 2.8565% (inside 0 2.6351%, outside both 0.2214%) | with CAL1 cluster 2.8489% F1 N=4 readout copy 0 fail 4.3495% (inside 0 4.1216%, outside both 0.2279%) | with CAL1 cluster 4.3486% | among shots verified 0 on every mid read 3.8323% F1 N=4 readout copy 1 fail 4.2788% (inside 0 4.0622%, outside both 0.2167%) | with CAL1 cluster 4.2779% | among shots verified 0 on every mid read 3.7993% F1 N=4 readout pooled fail 4.3142% (inside 0 4.0919%, outside both 0.2223%) | with CAL1 cluster 4.3133% | among shots verified 0 on every mid read 3.8158% F1 N=8 idle copy 0 fail 2.8950% (inside 0 2.6896%, outside both 0.2054%) | with CAL1 cluster 2.8901% F1 N=8 idle copy 1 fail 2.8712% (inside 0 2.6604%, outside both 0.2108%) | with CAL1 cluster 2.8653% F1 N=8 idle pooled fail 2.8831% (inside 0 2.6750%, outside both 0.2081%) | with CAL1 cluster 2.8777% F1 N=8 readout copy 0 fail 4.3532% (inside 0 4.1270%, outside both 0.2261%) | with CAL1 cluster 4.3509% | among shots verified 0 on every mid read 3.8103% F1 N=8 readout copy 1 fail 4.2905% (inside 0 4.0658%, outside both 0.2248%) | with CAL1 cluster 4.2910% | among shots verified 0 on every mid read 3.7410% F1 N=8 readout pooled fail 4.3218% (inside 0 4.0964%, outside both 0.2255%) | with CAL1 cluster 4.3209% | among shots verified 0 on every mid read 3.7756% KNOWN-GOOD idle N=0: flip failure minus CAL 0X floor -0.0353% [-0.1237, +0.0383] -> CONSISTENT -> PASS KNOWN-GOOD idle N=4: flip failure minus CAL 0X floor -0.0443% [-0.1318, +0.0406] -> CONSISTENT -> PASS KNOWN-GOOD idle N=8: flip failure minus CAL 0X floor -0.0177% [-0.0933, +0.0589] -> CONSISTENT -> PASS N=4 F2 hidden-leak rate (readout, verified-0 shots) - idle copy 0 +1.0237% [+0.9216, +1.1304] n=37 -> ABOVE N=4 F2 hidden-leak rate (readout, verified-0 shots) - idle copy 1 +0.8948% [+0.7946, +1.0019] n=37 -> ABOVE N=4 F2 hidden-leak rate (readout, verified-0 shots) - idle pooled +0.9592% [+0.8890, +1.0370] n=37 -> ABOVE N=4 F2 FINAL -> ABOVE N=4 F2u all-shot flip failure, readout - idle copy 0 +1.5410% [+1.4365, +1.6491] n=37 -> ABOVE N=4 F2u all-shot flip failure, readout - idle copy 1 +1.3743% [+1.2788, +1.4743] n=37 -> ABOVE N=4 F2u all-shot flip failure, readout - idle pooled +1.4577% [+1.3878, +1.5302] n=37 -> ABOVE N=4 F2u FINAL -> ABOVE N=4 L1w loss, mid-verified-0, ALL wired (run_iq R1 repeat) copy 0 +0.0121 [+0.0090, +0.0149] n=86 -> ABOVE N=4 L1w loss, mid-verified-0, ALL wired (run_iq R1 repeat) copy 1 +0.0117 [+0.0089, +0.0137] n=86 -> ABOVE N=4 L1w loss, mid-verified-0, ALL wired (run_iq R1 repeat) pooled +0.0095 [+0.0071, +0.0120] n=86 -> ABOVE N=4 L1w FINAL -> ABOVE N=4 L1 loss, mid-verified-0 selection copy 0 +0.0122 [+0.0090, +0.0151] n=83 -> ABOVE N=4 L1 loss, mid-verified-0 selection copy 1 +0.0111 [+0.0088, +0.0136] n=83 -> ABOVE N=4 L1 loss, mid-verified-0 selection pooled +0.0085 [+0.0071, +0.0119] n=83 -> ABOVE N=4 L1 FINAL -> ABOVE N=4 L2 loss, mid-verified-0 AND flipped copy 0 +0.0050 [+0.0007, +0.0074] n=83 -> ABOVE N=4 L2 loss, mid-verified-0 AND flipped copy 1 +0.0034 [+0.0016, +0.0063] n=83 -> ABOVE N=4 L2 loss, mid-verified-0 AND flipped pooled +0.0018 [-0.0003, +0.0049] n=83 -> CONSISTENT N=4 L2 FINAL -> CONSISTENT N=4 L3 L2 - L1 copy 0 -0.0076 [-0.0084, -0.0066] n=83 -> BELOW N=4 L3 L2 - L1 copy 1 -0.0075 [-0.0089, -0.0064] n=83 -> BELOW N=4 L3 L2 - L1 pooled -0.0077 [-0.0083, -0.0071] n=83 -> BELOW N=4 L3 FINAL -> BELOW KNOWN-GOOD N=4: L1w +0.0095 vs 21 Oct clean-shot loss +0.0087 [+0.0032, +0.0160] -> PASS N=4 --- marrakesh without neighbours of overlap/miss qubits (printed beside, does not replace): spectators used 81 (L1/L2/L3), 81 (L1w); dropped neighbours of [82, 100, 119] N=4 SIDE L1w copy 0 +0.0122 [+0.0090, +0.0154] n=81 -> ABOVE N=4 SIDE L1w copy 1 +0.0111 [+0.0089, +0.0136] n=81 -> ABOVE N=4 SIDE L1w pooled +0.0085 [+0.0071, +0.0120] n=81 -> ABOVE N=4 SIDE L1w FINAL -> ABOVE N=4 SIDE L1 copy 0 +0.0122 [+0.0090, +0.0154] n=81 -> ABOVE N=4 SIDE L1 copy 1 +0.0111 [+0.0088, +0.0136] n=81 -> ABOVE N=4 SIDE L1 pooled +0.0085 [+0.0071, +0.0120] n=81 -> ABOVE N=4 SIDE L1 FINAL -> ABOVE N=4 SIDE L2 copy 0 +0.0050 [+0.0005, +0.0075] n=81 -> ABOVE N=4 SIDE L2 copy 1 +0.0034 [+0.0016, +0.0063] n=81 -> ABOVE N=4 SIDE L2 pooled +0.0018 [-0.0003, +0.0049] n=81 -> CONSISTENT N=4 SIDE L2 FINAL -> CONSISTENT N=4 SIDE L3 copy 0 -0.0076 [-0.0084, -0.0066] n=81 -> BELOW N=4 SIDE L3 copy 1 -0.0075 [-0.0089, -0.0064] n=81 -> BELOW N=4 SIDE L3 pooled -0.0077 [-0.0083, -0.0071] n=81 -> BELOW N=4 SIDE L3 FINAL -> BELOW KNOWN-GOOD SIDE N=4: L1w without those neighbours +0.0085 vs 21 Oct clean-shot loss +0.0087 [+0.0032, +0.0160] -> PASS N=4 shots kept (median over wired spectators, pooled): L1 readout 98.6%, L2 readout 95.7%, L2 idle 98.1% N=4 DECISION -> L2 CONSISTENT while L1 ABOVE: the loss MAY come from neighbours stuck in a level that reads as 0 | L3 BELOW, F2 ABOVE N=8 F2 hidden-leak rate (readout, verified-0 shots) - idle copy 0 +0.9153% [+0.8179, +1.0138] n=37 -> ABOVE N=8 F2 hidden-leak rate (readout, verified-0 shots) - idle copy 1 +0.8699% [+0.7737, +0.9667] n=37 -> ABOVE N=8 F2 hidden-leak rate (readout, verified-0 shots) - idle pooled +0.8925% [+0.8198, +0.9643] n=37 -> ABOVE N=8 F2 FINAL -> ABOVE N=8 F2u all-shot flip failure, readout - idle copy 0 +1.4581% [+1.3563, +1.5595] n=37 -> ABOVE N=8 F2u all-shot flip failure, readout - idle copy 1 +1.4194% [+1.3045, +1.5347] n=37 -> ABOVE N=8 F2u all-shot flip failure, readout - idle pooled +1.4387% [+1.3612, +1.5196] n=37 -> ABOVE N=8 F2u FINAL -> ABOVE N=8 L1w loss, mid-verified-0, ALL wired (run_iq R1 repeat) copy 0 +0.0090 [+0.0033, +0.0168] n=86 -> ABOVE N=8 L1w loss, mid-verified-0, ALL wired (run_iq R1 repeat) copy 1 +0.0080 [+0.0011, +0.0125] n=86 -> ABOVE N=8 L1w loss, mid-verified-0, ALL wired (run_iq R1 repeat) pooled +0.0083 [+0.0051, +0.0112] n=86 -> ABOVE N=8 L1w FINAL -> ABOVE N=8 L1 loss, mid-verified-0 selection copy 0 +0.0094 [+0.0051, +0.0168] n=83 -> ABOVE N=8 L1 loss, mid-verified-0 selection copy 1 +0.0070 [-0.0002, +0.0122] n=83 -> CONSISTENT N=8 L1 loss, mid-verified-0 selection pooled +0.0085 [+0.0049, +0.0112] n=83 -> ABOVE N=8 L1 FINAL -> CANNOT_TELL (copies disagree) N=8 L2 loss, mid-verified-0 AND flipped copy 0 +0.0019 [-0.0040, +0.0092] n=83 -> CONSISTENT N=8 L2 loss, mid-verified-0 AND flipped copy 1 -0.0005 [-0.0082, +0.0051] n=83 -> CONSISTENT N=8 L2 loss, mid-verified-0 AND flipped pooled +0.0016 [-0.0027, +0.0060] n=83 -> CONSISTENT N=8 L2 FINAL -> CONSISTENT N=8 L3 L2 - L1 copy 0 -0.0067 [-0.0081, -0.0055] n=83 -> BELOW N=8 L3 L2 - L1 copy 1 -0.0070 [-0.0081, -0.0059] n=83 -> BELOW N=8 L3 L2 - L1 pooled -0.0072 [-0.0076, -0.0065] n=83 -> BELOW N=8 L3 FINAL -> BELOW KNOWN-GOOD N=8: L1w +0.0083 vs 21 Oct clean-shot loss +0.0205 [+0.0112, +0.0284] -> FAIL N=8 --- marrakesh without neighbours of overlap/miss qubits (printed beside, does not replace): spectators used 81 (L1/L2/L3), 81 (L1w); dropped neighbours of [82, 100, 119] N=8 SIDE L1w copy 0 +0.0086 [+0.0025, +0.0168] n=81 -> ABOVE N=8 SIDE L1w copy 1 +0.0064 [-0.0002, +0.0125] n=81 -> CONSISTENT N=8 SIDE L1w pooled +0.0081 [+0.0048, +0.0110] n=81 -> ABOVE N=8 SIDE L1w FINAL -> CANNOT_TELL (copies disagree) N=8 SIDE L1 copy 0 +0.0086 [+0.0025, +0.0168] n=81 -> ABOVE N=8 SIDE L1 copy 1 +0.0064 [-0.0008, +0.0125] n=81 -> CONSISTENT N=8 SIDE L1 pooled +0.0081 [+0.0049, +0.0109] n=81 -> ABOVE N=8 SIDE L1 FINAL -> CANNOT_TELL (copies disagree) N=8 SIDE L2 copy 0 +0.0011 [-0.0043, +0.0089] n=81 -> CONSISTENT N=8 SIDE L2 copy 1 -0.0017 [-0.0095, +0.0046] n=81 -> CONSISTENT N=8 SIDE L2 pooled +0.0007 [-0.0041, +0.0053] n=81 -> CONSISTENT N=8 SIDE L2 FINAL -> CONSISTENT N=8 SIDE L3 copy 0 -0.0070 [-0.0082, -0.0055] n=81 -> BELOW N=8 SIDE L3 copy 1 -0.0070 [-0.0081, -0.0059] n=81 -> BELOW N=8 SIDE L3 pooled -0.0072 [-0.0076, -0.0066] n=81 -> BELOW N=8 SIDE L3 FINAL -> BELOW KNOWN-GOOD SIDE N=8: L1w without those neighbours +0.0081 vs 21 Oct clean-shot loss +0.0205 [+0.0112, +0.0284] -> FAIL N=8 shots kept (median over wired spectators, pooled): L1 readout 97.6%, L2 readout 94.7%, L2 idle 98.1% N=8 DECISION -> L1 CANNOT_TELL (copies disagree), L2 CONSISTENT: neither branch of the rule; reported as is | L3 BELOW, F2 ABOVE ############################## SCENARIO null0 seed 20261123 ==================================================================================================== ibm_marrakesh job SYNTH-null-20261123 CAL 0X flip-rule failure per measured qubit: median 1.95%, max 98.59%; flip-unreadable (> 20%): [82, 100]; wired spectators 86, used for L1/L2/L3 83 KNOWN-GOOD 0X: flip success, median over measured qubits 98.05% (flip-readable 98.13%) -> PASS F1 N=0 idle copy 0 fail 2.9851% (inside 0 2.7586%, outside both 0.2266%) | with CAL1 cluster 2.9865% F1 N=0 idle copy 1 fail 2.9455% (inside 0 2.7338%, outside both 0.2117%) | with CAL1 cluster 2.9428% F1 N=0 idle pooled fail 2.9653% (inside 0 2.7462%, outside both 0.2191%) | with CAL1 cluster 2.9646% F1 N=4 idle copy 0 fail 2.9194% (inside 0 2.7207%, outside both 0.1986%) | with CAL1 cluster 2.9266% F1 N=4 idle copy 1 fail 2.9518% (inside 0 2.7225%, outside both 0.2293%) | with CAL1 cluster 2.9468% F1 N=4 idle pooled fail 2.9356% (inside 0 2.7216%, outside both 0.2140%) | with CAL1 cluster 2.9367% F1 N=4 readout copy 0 fail 3.4689% (inside 0 3.2608%, outside both 0.2081%) | with CAL1 cluster 3.4734% | among shots verified 0 on every mid read 2.9851% F1 N=4 readout copy 1 fail 3.5275% (inside 0 3.3117%, outside both 0.2158%) | with CAL1 cluster 3.5270% | among shots verified 0 on every mid read 3.0109% F1 N=4 readout pooled fail 3.4982% (inside 0 3.2863%, outside both 0.2119%) | with CAL1 cluster 3.5002% | among shots verified 0 on every mid read 2.9980% F1 N=8 idle copy 0 fail 2.9811% (inside 0 2.7698%, outside both 0.2113%) | with CAL1 cluster 2.9860% F1 N=8 idle copy 1 fail 2.8797% (inside 0 2.6482%, outside both 0.2315%) | with CAL1 cluster 2.8856% F1 N=8 idle pooled fail 2.9304% (inside 0 2.7090%, outside both 0.2214%) | with CAL1 cluster 2.9358% F1 N=8 readout copy 0 fail 3.4216% (inside 0 3.1977%, outside both 0.2239%) | with CAL1 cluster 3.4270% | among shots verified 0 on every mid read 2.9341% F1 N=8 readout copy 1 fail 3.4455% (inside 0 3.2360%, outside both 0.2095%) | with CAL1 cluster 3.4491% | among shots verified 0 on every mid read 2.9776% F1 N=8 readout pooled fail 3.4336% (inside 0 3.2169%, outside both 0.2167%) | with CAL1 cluster 3.4381% | among shots verified 0 on every mid read 2.9558% KNOWN-GOOD idle N=0: flip failure minus CAL 0X floor -0.0331% [-0.1096, +0.0402] -> CONSISTENT -> PASS KNOWN-GOOD idle N=4: flip failure minus CAL 0X floor -0.0628% [-0.1583, +0.0220] -> CONSISTENT -> PASS KNOWN-GOOD idle N=8: flip failure minus CAL 0X floor -0.0680% [-0.1551, +0.0166] -> CANNOT_TELL (copies disagree) -> PASS N=4 F2 hidden-leak rate (readout, verified-0 shots) - idle copy 0 +0.0657% [-0.0450, +0.1642] n=37 -> CONSISTENT N=4 F2 hidden-leak rate (readout, verified-0 shots) - idle copy 1 +0.0591% [-0.0664, +0.1767] n=37 -> CONSISTENT N=4 F2 hidden-leak rate (readout, verified-0 shots) - idle pooled +0.0624% [-0.0145, +0.1439] n=37 -> CONSISTENT N=4 F2 FINAL -> CONSISTENT N=4 F2u all-shot flip failure, readout - idle copy 0 +0.5495% [+0.4441, +0.6532] n=37 -> ABOVE N=4 F2u all-shot flip failure, readout - idle copy 1 +0.5757% [+0.4468, +0.6955] n=37 -> ABOVE N=4 F2u all-shot flip failure, readout - idle pooled +0.5626% [+0.4835, +0.6464] n=37 -> ABOVE N=4 F2u FINAL -> ABOVE N=4 L1w loss, mid-verified-0, ALL wired (run_iq R1 repeat) copy 0 -0.0008 [-0.0057, +0.0018] n=86 -> CONSISTENT N=4 L1w loss, mid-verified-0, ALL wired (run_iq R1 repeat) copy 1 +0.0005 [-0.0038, +0.0061] n=86 -> CONSISTENT N=4 L1w loss, mid-verified-0, ALL wired (run_iq R1 repeat) pooled -0.0011 [-0.0041, +0.0018] n=86 -> CONSISTENT N=4 L1w FINAL -> CONSISTENT N=4 L1 loss, mid-verified-0 selection copy 0 -0.0011 [-0.0071, +0.0009] n=83 -> CONSISTENT N=4 L1 loss, mid-verified-0 selection copy 1 +0.0002 [-0.0046, +0.0061] n=83 -> CONSISTENT N=4 L1 loss, mid-verified-0 selection pooled -0.0016 [-0.0043, +0.0016] n=83 -> CONSISTENT N=4 L1 FINAL -> CONSISTENT N=4 L2 loss, mid-verified-0 AND flipped copy 0 -0.0006 [-0.0068, +0.0014] n=83 -> CONSISTENT N=4 L2 loss, mid-verified-0 AND flipped copy 1 +0.0008 [-0.0034, +0.0050] n=83 -> CONSISTENT N=4 L2 loss, mid-verified-0 AND flipped pooled -0.0019 [-0.0049, +0.0009] n=83 -> CONSISTENT N=4 L2 FINAL -> CONSISTENT N=4 L3 L2 - L1 copy 0 -0.0003 [-0.0005, +0.0004] n=83 -> CONSISTENT N=4 L3 L2 - L1 copy 1 +0.0001 [-0.0006, +0.0009] n=83 -> CONSISTENT N=4 L3 L2 - L1 pooled -0.0001 [-0.0006, +0.0003] n=83 -> CONSISTENT N=4 L3 FINAL -> CONSISTENT KNOWN-GOOD N=4: L1w -0.0011 vs 21 Oct clean-shot loss +0.0087 [+0.0032, +0.0160] -> FAIL N=4 --- marrakesh without neighbours of overlap/miss qubits (printed beside, does not replace): spectators used 81 (L1/L2/L3), 81 (L1w); dropped neighbours of [82, 100, 119] N=4 SIDE L1w copy 0 -0.0009 [-0.0066, +0.0014] n=81 -> CONSISTENT N=4 SIDE L1w copy 1 -0.0001 [-0.0050, +0.0057] n=81 -> CONSISTENT N=4 SIDE L1w pooled -0.0012 [-0.0043, +0.0018] n=81 -> CONSISTENT N=4 SIDE L1w FINAL -> CONSISTENT N=4 SIDE L1 copy 0 -0.0009 [-0.0048, +0.0009] n=81 -> CONSISTENT N=4 SIDE L1 copy 1 -0.0001 [-0.0046, +0.0061] n=81 -> CONSISTENT N=4 SIDE L1 pooled -0.0012 [-0.0043, +0.0018] n=81 -> CONSISTENT N=4 SIDE L1 FINAL -> CONSISTENT N=4 SIDE L2 copy 0 -0.0000 [-0.0049, +0.0014] n=81 -> CONSISTENT N=4 SIDE L2 copy 1 +0.0008 [-0.0038, +0.0052] n=81 -> CONSISTENT N=4 SIDE L2 pooled -0.0017 [-0.0049, +0.0009] n=81 -> CONSISTENT N=4 SIDE L2 FINAL -> CONSISTENT N=4 SIDE L3 copy 0 -0.0003 [-0.0005, +0.0004] n=81 -> CONSISTENT N=4 SIDE L3 copy 1 +0.0001 [-0.0006, +0.0010] n=81 -> CONSISTENT N=4 SIDE L3 pooled -0.0001 [-0.0005, +0.0003] n=81 -> CONSISTENT N=4 SIDE L3 FINAL -> CONSISTENT KNOWN-GOOD SIDE N=4: L1w without those neighbours -0.0012 vs 21 Oct clean-shot loss +0.0087 [+0.0032, +0.0160] -> FAIL N=4 shots kept (median over wired spectators, pooled): L1 readout 98.5%, L2 readout 96.6%, L2 idle 98.1% N=4 DECISION -> L1 CONSISTENT, L2 CONSISTENT: neither branch of the rule; reported as is | L3 CONSISTENT, F2 CONSISTENT N=8 F2 hidden-leak rate (readout, verified-0 shots) - idle copy 0 -0.0470% [-0.1276, +0.0355] n=37 -> CONSISTENT N=8 F2 hidden-leak rate (readout, verified-0 shots) - idle copy 1 +0.0979% [+0.0177, +0.1824] n=37 -> ABOVE N=8 F2 hidden-leak rate (readout, verified-0 shots) - idle pooled +0.0254% [-0.0221, +0.0755] n=37 -> CONSISTENT N=8 F2 FINAL -> CANNOT_TELL (copies disagree) N=8 F2u all-shot flip failure, readout - idle copy 0 +0.4405% [+0.3662, +0.5172] n=37 -> ABOVE N=8 F2u all-shot flip failure, readout - idle copy 1 +0.5658% [+0.4775, +0.6559] n=37 -> ABOVE N=8 F2u all-shot flip failure, readout - idle pooled +0.5032% [+0.4534, +0.5534] n=37 -> ABOVE N=8 F2u FINAL -> ABOVE N=8 L1w loss, mid-verified-0, ALL wired (run_iq R1 repeat) copy 0 +0.0026 [-0.0044, +0.0059] n=86 -> CONSISTENT N=8 L1w loss, mid-verified-0, ALL wired (run_iq R1 repeat) copy 1 -0.0015 [-0.0054, +0.0011] n=86 -> CONSISTENT N=8 L1w loss, mid-verified-0, ALL wired (run_iq R1 repeat) pooled +0.0003 [-0.0016, +0.0025] n=86 -> CONSISTENT N=8 L1w FINAL -> CONSISTENT N=8 L1 loss, mid-verified-0 selection copy 0 +0.0019 [-0.0047, +0.0060] n=83 -> CONSISTENT N=8 L1 loss, mid-verified-0 selection copy 1 -0.0016 [-0.0057, +0.0004] n=83 -> CONSISTENT N=8 L1 loss, mid-verified-0 selection pooled +0.0001 [-0.0017, +0.0020] n=83 -> CONSISTENT N=8 L1 FINAL -> CONSISTENT N=8 L2 loss, mid-verified-0 AND flipped copy 0 +0.0029 [-0.0053, +0.0072] n=83 -> CONSISTENT N=8 L2 loss, mid-verified-0 AND flipped copy 1 -0.0010 [-0.0064, +0.0011] n=83 -> CONSISTENT N=8 L2 loss, mid-verified-0 AND flipped pooled +0.0006 [-0.0015, +0.0014] n=83 -> CONSISTENT N=8 L2 FINAL -> CONSISTENT N=8 L3 L2 - L1 copy 0 -0.0002 [-0.0012, +0.0006] n=83 -> CONSISTENT N=8 L3 L2 - L1 copy 1 -0.0004 [-0.0011, +0.0006] n=83 -> CONSISTENT N=8 L3 L2 - L1 pooled -0.0005 [-0.0008, -0.0001] n=83 -> BELOW N=8 L3 FINAL -> BELOW KNOWN-GOOD N=8: L1w +0.0003 vs 21 Oct clean-shot loss +0.0205 [+0.0112, +0.0284] -> FAIL N=8 --- marrakesh without neighbours of overlap/miss qubits (printed beside, does not replace): spectators used 81 (L1/L2/L3), 81 (L1w); dropped neighbours of [82, 100, 119] N=8 SIDE L1w copy 0 +0.0032 [-0.0048, +0.0073] n=81 -> CONSISTENT N=8 SIDE L1w copy 1 -0.0016 [-0.0056, +0.0007] n=81 -> CONSISTENT N=8 SIDE L1w pooled +0.0006 [-0.0017, +0.0026] n=81 -> CONSISTENT N=8 SIDE L1w FINAL -> CONSISTENT N=8 SIDE L1 copy 0 +0.0032 [-0.0047, +0.0073] n=81 -> CONSISTENT N=8 SIDE L1 copy 1 -0.0016 [-0.0057, +0.0014] n=81 -> CONSISTENT N=8 SIDE L1 pooled +0.0006 [-0.0017, +0.0026] n=81 -> CONSISTENT N=8 SIDE L1 FINAL -> CONSISTENT N=8 SIDE L2 copy 0 +0.0046 [-0.0056, +0.0072] n=81 -> CONSISTENT N=8 SIDE L2 copy 1 -0.0010 [-0.0064, +0.0020] n=81 -> CONSISTENT N=8 SIDE L2 pooled +0.0006 [-0.0015, +0.0014] n=81 -> CONSISTENT N=8 SIDE L2 FINAL -> CONSISTENT N=8 SIDE L3 copy 0 -0.0002 [-0.0012, +0.0006] n=81 -> CONSISTENT N=8 SIDE L3 copy 1 -0.0004 [-0.0012, +0.0006] n=81 -> CONSISTENT N=8 SIDE L3 pooled -0.0005 [-0.0008, -0.0001] n=81 -> BELOW N=8 SIDE L3 FINAL -> BELOW KNOWN-GOOD SIDE N=8: L1w without those neighbours +0.0006 vs 21 Oct clean-shot loss +0.0205 [+0.0112, +0.0284] -> FAIL N=8 shots kept (median over wired spectators, pooled): L1 readout 97.7%, L2 readout 95.7%, L2 idle 98.1% N=8 DECISION -> L1 CONSISTENT, L2 CONSISTENT: neither branch of the rule; reported as is | L3 BELOW, F2 CANNOT_TELL (copies disagree)