gpt2
Find this model in the GPT model summary
gpt2 Model Summary Plots
gpt2 Model Selected Details
layer_type
N
M
Q
alpha
D
alpha-hat
log_SN
% Rand
num_traps
num_fingers
rank_loss
layer_id
2
EMBEDDING
1024
768
1.33
1.55
0.08
5.56
3.58
22.11
0
0
0
8
CONV1D
2304
768
3.00
2.55
0.06
8.42
3.30
78.06
0
0
0
9
CONV1D
768
768
1.00
2.76
0.04
7.86
2.85
64.61
0
0
2
14
CONV1D
3072
768
4.00
3.98
0.04
12.91
3.25
85.60
0
0
0
15
CONV1D
3072
768
4.00
3.29
0.06
9.21
2.80
83.99
12
0
0
21
CONV1D
2304
768
3.00
3.21
0.04
8.73
2.72
78.94
0
0
0
22
CONV1D
768
768
1.00
2.54
0.06
7.28
2.86
70.13
11
0
1
27
CONV1D
3072
768
4.00
3.42
0.03
10.85
3.17
84.57
0
0
0
28
CONV1D
3072
768
4.00
3.47
0.05
10.13
2.92
87.37
7
0
0
34
CONV1D
2304
768
3.00
3.52
0.04
9.84
2.79
82.62
0
0
0
35
CONV1D
768
768
1.00
3.58
0.04
8.19
2.29
83.18
0
0
0
40
CONV1D
3072
768
4.00
3.18
0.03
10.09
3.18
83.79
2
0
0
41
CONV1D
3072
768
4.00
3.32
0.02
10.51
3.17
85.78
6
0
0
47
CONV1D
2304
768
3.00
3.22
0.03
8.59
2.67
85.65
0
0
0
48
CONV1D
768
768
1.00
4.69
0.04
10.44
2.23
85.59
0
0
1
53
CONV1D
3072
768
4.00
3.64
0.02
11.01
3.02
86.56
1
0
0
54
CONV1D
3072
768
4.00
3.52
0.04
11.33
3.22
84.38
4
0
0
60
CONV1D
2304
768
3.00
2.98
0.05
7.89
2.65
84.06
0
0
0
61
CONV1D
768
768
1.00
2.94
0.11
6.83
2.32
84.59
0
1
0
66
CONV1D
3072
768
4.00
3.45
0.03
10.36
3.00
85.18
0
0
0
67
CONV1D
3072
768
4.00
3.60
0.03
10.81
3.00
83.54
2
0
0
73
CONV1D
2304
768
3.00
2.99
0.09
7.29
2.44
82.91
0
0
0
74
CONV1D
768
768
1.00
3.65
0.04
8.10
2.22
87.42
0
0
0
79
CONV1D
3072
768
4.00
3.90
0.03
11.79
3.02
85.49
1
0
0
80
CONV1D
3072
768
4.00
3.59
0.04
10.36
2.88
85.05
0
0
0
86
CONV1D
2304
768
3.00
3.14
0.08
7.48
2.38
85.88
0
0
0
87
CONV1D
768
768
1.00
3.27
0.11
7.68
2.35
84.49
0
1
1
92
CONV1D
3072
768
4.00
3.50
0.04
10.49
2.99
85.73
0
0
0
93
CONV1D
3072
768
4.00
3.87
0.03
11.41
2.94
85.95
0
0
0
99
CONV1D
2304
768
3.00
3.65
0.08
8.81
2.42
86.45
0
0
0
100
CONV1D
768
768
1.00
3.09
0.08
7.63
2.47
86.17
0
1
1
105
CONV1D
3072
768
4.00
3.74
0.04
11.24
3.00
86.01
0
0
0
106
CONV1D
3072
768
4.00
4.05
0.02
12.01
2.96
87.09
0
0
0
112
CONV1D
2304
768
3.00
4.64
0.07
10.97
2.36
90.31
0
0
0
113
CONV1D
768
768
1.00
2.89
0.13
7.53
2.61
85.45
0
3
1
118
CONV1D
3072
768
4.00
4.01
0.03
12.00
2.99
87.00
0
0
0
119
CONV1D
3072
768
4.00
4.18
0.02
12.46
2.98
87.19
0
0
0
125
CONV1D
2304
768
3.00
4.60
0.06
10.80
2.35
92.75
0
0
0
126
CONV1D
768
768
1.00
3.87
0.13
10.45
2.70
89.30
0
6
1
131
CONV1D
3072
768
4.00
3.67
0.05
11.02
3.00
86.75
0
0
0
132
CONV1D
3072
768
4.00
4.56
0.03
13.83
3.03
88.38
0
0
0
138
CONV1D
2304
768
3.00
6.53
0.05
15.35
2.35
90.09
0
0
0
139
CONV1D
768
768
1.00
4.96
0.09
15.02
3.03
85.02
1
0
0
144
CONV1D
3072
768
4.00
3.59
0.03
10.84
3.02
86.21
0
0
0
145
CONV1D
3072
768
4.00
4.53
0.02
14.66
3.24
89.11
2
0
0
151
CONV1D
2304
768
3.00
2.48
0.12
5.83
2.35
88.45
0
6
0
152
CONV1D
768
768
1.00
2.99
0.11
10.38
3.48
70.57
17
0
0
157
CONV1D
3072
768
4.00
3.32
0.03
10.44
3.15
84.31
0
0
0
158
CONV1D
3072
768
4.00
4.22
0.04
15.08
3.57
86.95
1
0
0
gpt2 Layer Plots
Layer 2
Layer=2 | N=1024 | M=768 | Q=1.33 | alpha=1.55 | D_ks=0.08 | alpha-hat=5.56 | num traps=0
Layer 8
Layer=8 | N=2304 | M=768 | Q=3.00 | alpha=2.55 | D_ks=0.06 | alpha-hat=8.42 | num traps=0
Layer 9
Layer=9 | N=768 | M=768 | Q=1.00 | alpha=2.76 | D_ks=0.04 | alpha-hat=7.86 | num traps=0
Layer 14
Layer=14 | N=3072 | M=768 | Q=4.00 | alpha=3.98 | D_ks=0.04 | alpha-hat=12.91 | num traps=0
Layer 15
Layer=15 | N=3072 | M=768 | Q=4.00 | alpha=3.29 | D_ks=0.06 | alpha-hat=9.21 | num traps=12
Layer 21
Layer=21 | N=2304 | M=768 | Q=3.00 | alpha=3.21 | D_ks=0.04 | alpha-hat=8.73 | num traps=0
Layer 22
Layer=22 | N=768 | M=768 | Q=1.00 | alpha=2.54 | D_ks=0.06 | alpha-hat=7.28 | num traps=11
Layer 27
Layer=27 | N=3072 | M=768 | Q=4.00 | alpha=3.42 | D_ks=0.03 | alpha-hat=10.85 | num traps=0
Layer 28
Layer=28 | N=3072 | M=768 | Q=4.00 | alpha=3.47 | D_ks=0.05 | alpha-hat=10.13 | num traps=7
Layer 34
Layer=34 | N=2304 | M=768 | Q=3.00 | alpha=3.52 | D_ks=0.04 | alpha-hat=9.84 | num traps=0
Layer 35
Layer=35 | N=768 | M=768 | Q=1.00 | alpha=3.58 | D_ks=0.04 | alpha-hat=8.19 | num traps=0
Layer 40
Layer=40 | N=3072 | M=768 | Q=4.00 | alpha=3.18 | D_ks=0.03 | alpha-hat=10.09 | num traps=2
Layer 41
Layer=41 | N=3072 | M=768 | Q=4.00 | alpha=3.32 | D_ks=0.02 | alpha-hat=10.51 | num traps=6
Layer 47
Layer=47 | N=2304 | M=768 | Q=3.00 | alpha=3.22 | D_ks=0.03 | alpha-hat=8.59 | num traps=0
Layer 48
Layer=48 | N=768 | M=768 | Q=1.00 | alpha=4.69 | D_ks=0.04 | alpha-hat=10.44 | num traps=0
Layer 53
Layer=53 | N=3072 | M=768 | Q=4.00 | alpha=3.64 | D_ks=0.02 | alpha-hat=11.01 | num traps=1
Layer 54
Layer=54 | N=3072 | M=768 | Q=4.00 | alpha=3.52 | D_ks=0.04 | alpha-hat=11.33 | num traps=4
Layer 60
Layer=60 | N=2304 | M=768 | Q=3.00 | alpha=2.98 | D_ks=0.05 | alpha-hat=7.89 | num traps=0
Layer 61
Layer=61 | N=768 | M=768 | Q=1.00 | alpha=2.94 | D_ks=0.11 | alpha-hat=6.83 | num traps=0
Layer 66
Layer=66 | N=3072 | M=768 | Q=4.00 | alpha=3.45 | D_ks=0.03 | alpha-hat=10.36 | num traps=0
Layer 67
Layer=67 | N=3072 | M=768 | Q=4.00 | alpha=3.60 | D_ks=0.03 | alpha-hat=10.81 | num traps=2
Layer 73
Layer=73 | N=2304 | M=768 | Q=3.00 | alpha=2.99 | D_ks=0.09 | alpha-hat=7.29 | num traps=0
Layer 74
Layer=74 | N=768 | M=768 | Q=1.00 | alpha=3.65 | D_ks=0.04 | alpha-hat=8.10 | num traps=0
Layer 79
Layer=79 | N=3072 | M=768 | Q=4.00 | alpha=3.90 | D_ks=0.03 | alpha-hat=11.79 | num traps=1
Layer 80
Layer=80 | N=3072 | M=768 | Q=4.00 | alpha=3.59 | D_ks=0.04 | alpha-hat=10.36 | num traps=0
Layer 86
Layer=86 | N=2304 | M=768 | Q=3.00 | alpha=3.14 | D_ks=0.08 | alpha-hat=7.48 | num traps=0
Layer 87
Layer=87 | N=768 | M=768 | Q=1.00 | alpha=3.27 | D_ks=0.11 | alpha-hat=7.68 | num traps=0
Layer 92
Layer=92 | N=3072 | M=768 | Q=4.00 | alpha=3.50 | D_ks=0.04 | alpha-hat=10.49 | num traps=0
Layer 93
Layer=93 | N=3072 | M=768 | Q=4.00 | alpha=3.87 | D_ks=0.03 | alpha-hat=11.41 | num traps=0
Layer 99
Layer=99 | N=2304 | M=768 | Q=3.00 | alpha=3.65 | D_ks=0.08 | alpha-hat=8.81 | num traps=0
Layer 100
Layer=100 | N=768 | M=768 | Q=1.00 | alpha=3.09 | D_ks=0.08 | alpha-hat=7.63 | num traps=0
Layer 105
Layer=105 | N=3072 | M=768 | Q=4.00 | alpha=3.74 | D_ks=0.04 | alpha-hat=11.24 | num traps=0
Layer 106
Layer=106 | N=3072 | M=768 | Q=4.00 | alpha=4.05 | D_ks=0.02 | alpha-hat=12.01 | num traps=0
Layer 112
Layer=112 | N=2304 | M=768 | Q=3.00 | alpha=4.64 | D_ks=0.07 | alpha-hat=10.97 | num traps=0
Layer 113
Layer=113 | N=768 | M=768 | Q=1.00 | alpha=2.89 | D_ks=0.13 | alpha-hat=7.53 | num traps=0
Layer 118
Layer=118 | N=3072 | M=768 | Q=4.00 | alpha=4.01 | D_ks=0.03 | alpha-hat=12.00 | num traps=0
Layer 119
Layer=119 | N=3072 | M=768 | Q=4.00 | alpha=4.18 | D_ks=0.02 | alpha-hat=12.46 | num traps=0
Layer 125
Layer=125 | N=2304 | M=768 | Q=3.00 | alpha=4.60 | D_ks=0.06 | alpha-hat=10.80 | num traps=0
Layer 126
Layer=126 | N=768 | M=768 | Q=1.00 | alpha=3.87 | D_ks=0.13 | alpha-hat=10.45 | num traps=0
Layer 131
Layer=131 | N=3072 | M=768 | Q=4.00 | alpha=3.67 | D_ks=0.05 | alpha-hat=11.02 | num traps=0
Layer 132
Layer=132 | N=3072 | M=768 | Q=4.00 | alpha=4.56 | D_ks=0.03 | alpha-hat=13.83 | num traps=0
Layer 138
Layer=138 | N=2304 | M=768 | Q=3.00 | alpha=6.53 | D_ks=0.05 | alpha-hat=15.35 | num traps=0
Layer 139
Layer=139 | N=768 | M=768 | Q=1.00 | alpha=4.96 | D_ks=0.09 | alpha-hat=15.02 | num traps=1
Layer 144
Layer=144 | N=3072 | M=768 | Q=4.00 | alpha=3.59 | D_ks=0.03 | alpha-hat=10.84 | num traps=0
Layer 145
Layer=145 | N=3072 | M=768 | Q=4.00 | alpha=4.53 | D_ks=0.02 | alpha-hat=14.66 | num traps=2
Layer 151
Layer=151 | N=2304 | M=768 | Q=3.00 | alpha=2.48 | D_ks=0.12 | alpha-hat=5.83 | num traps=0
Layer 152
Layer=152 | N=768 | M=768 | Q=1.00 | alpha=2.99 | D_ks=0.11 | alpha-hat=10.38 | num traps=17
Layer 157
Layer=157 | N=3072 | M=768 | Q=4.00 | alpha=3.32 | D_ks=0.03 | alpha-hat=10.44 | num traps=0
Layer 158
Layer=158 | N=3072 | M=768 | Q=4.00 | alpha=4.22 | D_ks=0.04 | alpha-hat=15.08 | num traps=1
gpt2 Model Summary Plots
gpt2 Model Selected Details
| layer_type | N | M | Q | alpha | D | alpha-hat | log_SN | % Rand | num_traps | num_fingers | rank_loss | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| layer_id | ||||||||||||
| 2 | EMBEDDING | 1024 | 768 | 1.33 | 1.55 | 0.08 | 5.56 | 3.58 | 22.11 | 0 | 0 | 0 |
| 8 | CONV1D | 2304 | 768 | 3.00 | 2.55 | 0.06 | 8.42 | 3.30 | 78.06 | 0 | 0 | 0 |
| 9 | CONV1D | 768 | 768 | 1.00 | 2.76 | 0.04 | 7.86 | 2.85 | 64.61 | 0 | 0 | 2 |
| 14 | CONV1D | 3072 | 768 | 4.00 | 3.98 | 0.04 | 12.91 | 3.25 | 85.60 | 0 | 0 | 0 |
| 15 | CONV1D | 3072 | 768 | 4.00 | 3.29 | 0.06 | 9.21 | 2.80 | 83.99 | 12 | 0 | 0 |
| 21 | CONV1D | 2304 | 768 | 3.00 | 3.21 | 0.04 | 8.73 | 2.72 | 78.94 | 0 | 0 | 0 |
| 22 | CONV1D | 768 | 768 | 1.00 | 2.54 | 0.06 | 7.28 | 2.86 | 70.13 | 11 | 0 | 1 |
| 27 | CONV1D | 3072 | 768 | 4.00 | 3.42 | 0.03 | 10.85 | 3.17 | 84.57 | 0 | 0 | 0 |
| 28 | CONV1D | 3072 | 768 | 4.00 | 3.47 | 0.05 | 10.13 | 2.92 | 87.37 | 7 | 0 | 0 |
| 34 | CONV1D | 2304 | 768 | 3.00 | 3.52 | 0.04 | 9.84 | 2.79 | 82.62 | 0 | 0 | 0 |
| 35 | CONV1D | 768 | 768 | 1.00 | 3.58 | 0.04 | 8.19 | 2.29 | 83.18 | 0 | 0 | 0 |
| 40 | CONV1D | 3072 | 768 | 4.00 | 3.18 | 0.03 | 10.09 | 3.18 | 83.79 | 2 | 0 | 0 |
| 41 | CONV1D | 3072 | 768 | 4.00 | 3.32 | 0.02 | 10.51 | 3.17 | 85.78 | 6 | 0 | 0 |
| 47 | CONV1D | 2304 | 768 | 3.00 | 3.22 | 0.03 | 8.59 | 2.67 | 85.65 | 0 | 0 | 0 |
| 48 | CONV1D | 768 | 768 | 1.00 | 4.69 | 0.04 | 10.44 | 2.23 | 85.59 | 0 | 0 | 1 |
| 53 | CONV1D | 3072 | 768 | 4.00 | 3.64 | 0.02 | 11.01 | 3.02 | 86.56 | 1 | 0 | 0 |
| 54 | CONV1D | 3072 | 768 | 4.00 | 3.52 | 0.04 | 11.33 | 3.22 | 84.38 | 4 | 0 | 0 |
| 60 | CONV1D | 2304 | 768 | 3.00 | 2.98 | 0.05 | 7.89 | 2.65 | 84.06 | 0 | 0 | 0 |
| 61 | CONV1D | 768 | 768 | 1.00 | 2.94 | 0.11 | 6.83 | 2.32 | 84.59 | 0 | 1 | 0 |
| 66 | CONV1D | 3072 | 768 | 4.00 | 3.45 | 0.03 | 10.36 | 3.00 | 85.18 | 0 | 0 | 0 |
| 67 | CONV1D | 3072 | 768 | 4.00 | 3.60 | 0.03 | 10.81 | 3.00 | 83.54 | 2 | 0 | 0 |
| 73 | CONV1D | 2304 | 768 | 3.00 | 2.99 | 0.09 | 7.29 | 2.44 | 82.91 | 0 | 0 | 0 |
| 74 | CONV1D | 768 | 768 | 1.00 | 3.65 | 0.04 | 8.10 | 2.22 | 87.42 | 0 | 0 | 0 |
| 79 | CONV1D | 3072 | 768 | 4.00 | 3.90 | 0.03 | 11.79 | 3.02 | 85.49 | 1 | 0 | 0 |
| 80 | CONV1D | 3072 | 768 | 4.00 | 3.59 | 0.04 | 10.36 | 2.88 | 85.05 | 0 | 0 | 0 |
| 86 | CONV1D | 2304 | 768 | 3.00 | 3.14 | 0.08 | 7.48 | 2.38 | 85.88 | 0 | 0 | 0 |
| 87 | CONV1D | 768 | 768 | 1.00 | 3.27 | 0.11 | 7.68 | 2.35 | 84.49 | 0 | 1 | 1 |
| 92 | CONV1D | 3072 | 768 | 4.00 | 3.50 | 0.04 | 10.49 | 2.99 | 85.73 | 0 | 0 | 0 |
| 93 | CONV1D | 3072 | 768 | 4.00 | 3.87 | 0.03 | 11.41 | 2.94 | 85.95 | 0 | 0 | 0 |
| 99 | CONV1D | 2304 | 768 | 3.00 | 3.65 | 0.08 | 8.81 | 2.42 | 86.45 | 0 | 0 | 0 |
| 100 | CONV1D | 768 | 768 | 1.00 | 3.09 | 0.08 | 7.63 | 2.47 | 86.17 | 0 | 1 | 1 |
| 105 | CONV1D | 3072 | 768 | 4.00 | 3.74 | 0.04 | 11.24 | 3.00 | 86.01 | 0 | 0 | 0 |
| 106 | CONV1D | 3072 | 768 | 4.00 | 4.05 | 0.02 | 12.01 | 2.96 | 87.09 | 0 | 0 | 0 |
| 112 | CONV1D | 2304 | 768 | 3.00 | 4.64 | 0.07 | 10.97 | 2.36 | 90.31 | 0 | 0 | 0 |
| 113 | CONV1D | 768 | 768 | 1.00 | 2.89 | 0.13 | 7.53 | 2.61 | 85.45 | 0 | 3 | 1 |
| 118 | CONV1D | 3072 | 768 | 4.00 | 4.01 | 0.03 | 12.00 | 2.99 | 87.00 | 0 | 0 | 0 |
| 119 | CONV1D | 3072 | 768 | 4.00 | 4.18 | 0.02 | 12.46 | 2.98 | 87.19 | 0 | 0 | 0 |
| 125 | CONV1D | 2304 | 768 | 3.00 | 4.60 | 0.06 | 10.80 | 2.35 | 92.75 | 0 | 0 | 0 |
| 126 | CONV1D | 768 | 768 | 1.00 | 3.87 | 0.13 | 10.45 | 2.70 | 89.30 | 0 | 6 | 1 |
| 131 | CONV1D | 3072 | 768 | 4.00 | 3.67 | 0.05 | 11.02 | 3.00 | 86.75 | 0 | 0 | 0 |
| 132 | CONV1D | 3072 | 768 | 4.00 | 4.56 | 0.03 | 13.83 | 3.03 | 88.38 | 0 | 0 | 0 |
| 138 | CONV1D | 2304 | 768 | 3.00 | 6.53 | 0.05 | 15.35 | 2.35 | 90.09 | 0 | 0 | 0 |
| 139 | CONV1D | 768 | 768 | 1.00 | 4.96 | 0.09 | 15.02 | 3.03 | 85.02 | 1 | 0 | 0 |
| 144 | CONV1D | 3072 | 768 | 4.00 | 3.59 | 0.03 | 10.84 | 3.02 | 86.21 | 0 | 0 | 0 |
| 145 | CONV1D | 3072 | 768 | 4.00 | 4.53 | 0.02 | 14.66 | 3.24 | 89.11 | 2 | 0 | 0 |
| 151 | CONV1D | 2304 | 768 | 3.00 | 2.48 | 0.12 | 5.83 | 2.35 | 88.45 | 0 | 6 | 0 |
| 152 | CONV1D | 768 | 768 | 1.00 | 2.99 | 0.11 | 10.38 | 3.48 | 70.57 | 17 | 0 | 0 |
| 157 | CONV1D | 3072 | 768 | 4.00 | 3.32 | 0.03 | 10.44 | 3.15 | 84.31 | 0 | 0 | 0 |
| 158 | CONV1D | 3072 | 768 | 4.00 | 4.22 | 0.04 | 15.08 | 3.57 | 86.95 | 1 | 0 | 0 |
gpt2 Layer Plots
Layer 2
Layer=2 | N=1024 | M=768 | Q=1.33 | alpha=1.55 | D_ks=0.08 | alpha-hat=5.56 | num traps=0
Layer 8
Layer=8 | N=2304 | M=768 | Q=3.00 | alpha=2.55 | D_ks=0.06 | alpha-hat=8.42 | num traps=0
Layer 9
Layer=9 | N=768 | M=768 | Q=1.00 | alpha=2.76 | D_ks=0.04 | alpha-hat=7.86 | num traps=0
Layer 14
Layer=14 | N=3072 | M=768 | Q=4.00 | alpha=3.98 | D_ks=0.04 | alpha-hat=12.91 | num traps=0
Layer 15
Layer=15 | N=3072 | M=768 | Q=4.00 | alpha=3.29 | D_ks=0.06 | alpha-hat=9.21 | num traps=12
Layer 21
Layer=21 | N=2304 | M=768 | Q=3.00 | alpha=3.21 | D_ks=0.04 | alpha-hat=8.73 | num traps=0
Layer 22
Layer=22 | N=768 | M=768 | Q=1.00 | alpha=2.54 | D_ks=0.06 | alpha-hat=7.28 | num traps=11
Layer 27
Layer=27 | N=3072 | M=768 | Q=4.00 | alpha=3.42 | D_ks=0.03 | alpha-hat=10.85 | num traps=0
Layer 28
Layer=28 | N=3072 | M=768 | Q=4.00 | alpha=3.47 | D_ks=0.05 | alpha-hat=10.13 | num traps=7
Layer 34
Layer=34 | N=2304 | M=768 | Q=3.00 | alpha=3.52 | D_ks=0.04 | alpha-hat=9.84 | num traps=0
Layer 35
Layer=35 | N=768 | M=768 | Q=1.00 | alpha=3.58 | D_ks=0.04 | alpha-hat=8.19 | num traps=0
Layer 40
Layer=40 | N=3072 | M=768 | Q=4.00 | alpha=3.18 | D_ks=0.03 | alpha-hat=10.09 | num traps=2
Layer 41
Layer=41 | N=3072 | M=768 | Q=4.00 | alpha=3.32 | D_ks=0.02 | alpha-hat=10.51 | num traps=6
Layer 47
Layer=47 | N=2304 | M=768 | Q=3.00 | alpha=3.22 | D_ks=0.03 | alpha-hat=8.59 | num traps=0
Layer 48
Layer=48 | N=768 | M=768 | Q=1.00 | alpha=4.69 | D_ks=0.04 | alpha-hat=10.44 | num traps=0
Layer 53
Layer=53 | N=3072 | M=768 | Q=4.00 | alpha=3.64 | D_ks=0.02 | alpha-hat=11.01 | num traps=1
Layer 54
Layer=54 | N=3072 | M=768 | Q=4.00 | alpha=3.52 | D_ks=0.04 | alpha-hat=11.33 | num traps=4
Layer 60
Layer=60 | N=2304 | M=768 | Q=3.00 | alpha=2.98 | D_ks=0.05 | alpha-hat=7.89 | num traps=0
Layer 61
Layer=61 | N=768 | M=768 | Q=1.00 | alpha=2.94 | D_ks=0.11 | alpha-hat=6.83 | num traps=0
Layer 66
Layer=66 | N=3072 | M=768 | Q=4.00 | alpha=3.45 | D_ks=0.03 | alpha-hat=10.36 | num traps=0
Layer 67
Layer=67 | N=3072 | M=768 | Q=4.00 | alpha=3.60 | D_ks=0.03 | alpha-hat=10.81 | num traps=2
Layer 73
Layer=73 | N=2304 | M=768 | Q=3.00 | alpha=2.99 | D_ks=0.09 | alpha-hat=7.29 | num traps=0
Layer 74
Layer=74 | N=768 | M=768 | Q=1.00 | alpha=3.65 | D_ks=0.04 | alpha-hat=8.10 | num traps=0
Layer 79
Layer=79 | N=3072 | M=768 | Q=4.00 | alpha=3.90 | D_ks=0.03 | alpha-hat=11.79 | num traps=1
Layer 80
Layer=80 | N=3072 | M=768 | Q=4.00 | alpha=3.59 | D_ks=0.04 | alpha-hat=10.36 | num traps=0
Layer 86
Layer=86 | N=2304 | M=768 | Q=3.00 | alpha=3.14 | D_ks=0.08 | alpha-hat=7.48 | num traps=0
Layer 87
Layer=87 | N=768 | M=768 | Q=1.00 | alpha=3.27 | D_ks=0.11 | alpha-hat=7.68 | num traps=0
Layer 92
Layer=92 | N=3072 | M=768 | Q=4.00 | alpha=3.50 | D_ks=0.04 | alpha-hat=10.49 | num traps=0
Layer 93
Layer=93 | N=3072 | M=768 | Q=4.00 | alpha=3.87 | D_ks=0.03 | alpha-hat=11.41 | num traps=0
Layer 99
Layer=99 | N=2304 | M=768 | Q=3.00 | alpha=3.65 | D_ks=0.08 | alpha-hat=8.81 | num traps=0
Layer 100
Layer=100 | N=768 | M=768 | Q=1.00 | alpha=3.09 | D_ks=0.08 | alpha-hat=7.63 | num traps=0
Layer 105
Layer=105 | N=3072 | M=768 | Q=4.00 | alpha=3.74 | D_ks=0.04 | alpha-hat=11.24 | num traps=0
Layer 106
Layer=106 | N=3072 | M=768 | Q=4.00 | alpha=4.05 | D_ks=0.02 | alpha-hat=12.01 | num traps=0
Layer 112
Layer=112 | N=2304 | M=768 | Q=3.00 | alpha=4.64 | D_ks=0.07 | alpha-hat=10.97 | num traps=0
Layer 113
Layer=113 | N=768 | M=768 | Q=1.00 | alpha=2.89 | D_ks=0.13 | alpha-hat=7.53 | num traps=0
Layer 118
Layer=118 | N=3072 | M=768 | Q=4.00 | alpha=4.01 | D_ks=0.03 | alpha-hat=12.00 | num traps=0
Layer 119
Layer=119 | N=3072 | M=768 | Q=4.00 | alpha=4.18 | D_ks=0.02 | alpha-hat=12.46 | num traps=0
Layer 125
Layer=125 | N=2304 | M=768 | Q=3.00 | alpha=4.60 | D_ks=0.06 | alpha-hat=10.80 | num traps=0
Layer 126
Layer=126 | N=768 | M=768 | Q=1.00 | alpha=3.87 | D_ks=0.13 | alpha-hat=10.45 | num traps=0
Layer 131
Layer=131 | N=3072 | M=768 | Q=4.00 | alpha=3.67 | D_ks=0.05 | alpha-hat=11.02 | num traps=0
Layer 132
Layer=132 | N=3072 | M=768 | Q=4.00 | alpha=4.56 | D_ks=0.03 | alpha-hat=13.83 | num traps=0
Layer 138
Layer=138 | N=2304 | M=768 | Q=3.00 | alpha=6.53 | D_ks=0.05 | alpha-hat=15.35 | num traps=0
Layer 139
Layer=139 | N=768 | M=768 | Q=1.00 | alpha=4.96 | D_ks=0.09 | alpha-hat=15.02 | num traps=1
Layer 144
Layer=144 | N=3072 | M=768 | Q=4.00 | alpha=3.59 | D_ks=0.03 | alpha-hat=10.84 | num traps=0
Layer 145
Layer=145 | N=3072 | M=768 | Q=4.00 | alpha=4.53 | D_ks=0.02 | alpha-hat=14.66 | num traps=2
Layer 151
Layer=151 | N=2304 | M=768 | Q=3.00 | alpha=2.48 | D_ks=0.12 | alpha-hat=5.83 | num traps=0
Layer 152
Layer=152 | N=768 | M=768 | Q=1.00 | alpha=2.99 | D_ks=0.11 | alpha-hat=10.38 | num traps=17
Layer 157
Layer=157 | N=3072 | M=768 | Q=4.00 | alpha=3.32 | D_ks=0.03 | alpha-hat=10.44 | num traps=0
Layer 158
Layer=158 | N=3072 | M=768 | Q=4.00 | alpha=4.22 | D_ks=0.04 | alpha-hat=15.08 | num traps=1