layer_id
int64
0
223
name
stringlengths
26
32
D
float64
0.03
0.12
M
int64
1.02k
4.1k
N
int64
4.1k
14.3k
Q
float64
1
4
alpha
float64
2.98
23.9
alpha_weighted
float64
-65.71
-6.41
entropy
float64
1.11
1.57
has_esd
bool
1 class
lambda_max
float32
0
0.02
layer_type
stringclasses
1 value
log_alpha_norm
float64
-64.96
-5.95
log_norm
float32
-1.43
-0.48
log_spectral_norm
float32
-2.81
-1.77
matrix_rank
int64
64
64
norm
float32
0.04
0.33
num_evals
int64
1.02k
4.1k
num_pl_spikes
int64
10
64
rank_loss
int64
960
4.03k
rf
int64
1
1
sigma
float64
0.25
5.88
spectral_norm
float32
0
0.02
stable_rank
float32
7.52
56.2
status
stringclasses
1 value
sv_max
float64
0.04
0.13
sv_min
float64
0
0
warning
stringclasses
2 values
weak_rank_loss
int64
960
4.03k
xmax
float64
0
0.02
xmin
float64
0
0
100
model.layers.14.mlp.up_proj
0.04407
4,096
14,336
3.5
5.35609
-9.927089
1.560365
true
0.014015
dense
-9.748356
-0.481377
-1.853421
64
0.330083
4,096
24
4,032
1
0.889183
0.014015
23.552853
success
0.118383
0.000001
4,032
0.014015
0.004832
101
model.layers.14.self_attn.k_proj
0.050633
1,024
4,096
4
7.090553
-17.852702
1.13429
true
0.003035
dense
-17.129008
-0.910717
-2.517815
64
0.122824
1,024
51
960
1
0.852848
0.003035
40.466755
success
0.055092
0.000001
under-trained
960
0.003035
0.001672
102
model.layers.14.self_attn.o_proj
0.027317
4,096
4,096
1
10.210347
-23.017842
1.566375
true
0.005567
dense
-22.823744
-0.681389
-2.254364
64
0.208262
4,096
56
4,032
1
1.230784
0.005567
37.408894
success
0.074614
0
under-trained
4,032
0.005567
0.002941
103
model.layers.14.self_attn.q_proj
0.049554
4,096
4,096
1
8.232165
-19.06172
1.564752
true
0.004836
dense
-18.963099
-0.828161
-2.315517
64
0.148538
4,096
33
4,032
1
1.258958
0.004836
30.715399
success
0.069541
0
under-trained
4,032
0.004836
0.002207
104
model.layers.14.self_attn.v_proj
0.079832
1,024
4,096
4
20.738459
-53.412723
1.136854
true
0.002657
dense
-53.051262
-0.87145
-2.57554
64
0.134447
1,024
17
960
1
4.78728
0.002657
50.592876
success
0.05155
0.000001
under-trained
960
0.002657
0.002202
105
model.layers.15.mlp.down_proj
0.029379
4,096
14,336
3.5
14.231365
-30.512898
1.567433
true
0.007177
dense
-30.400055
-0.516115
-2.14406
64
0.304709
4,096
64
4,032
1
1.653921
0.007177
42.456596
success
0.084717
0.000001
under-trained
4,032
0.007177
0.004398
106
model.layers.15.mlp.gate_proj
0.051791
4,096
14,336
3.5
5.879363
-11.277452
1.560675
true
0.012074
dense
-11.035696
-0.507647
-1.918142
64
0.310708
4,096
35
4,032
1
0.824763
0.012074
25.733248
success
0.109883
0.000001
4,032
0.012074
0.004324
107
model.layers.15.mlp.up_proj
0.04936
4,096
14,336
3.5
6.388651
-12.393745
1.562404
true
0.011483
dense
-12.144656
-0.49289
-1.939963
64
0.321448
4,096
30
4,032
1
0.983829
0.011483
27.99452
success
0.107157
0.000001
under-trained
4,032
0.011483
0.00465
108
model.layers.15.self_attn.k_proj
0.08746
1,024
4,096
4
12.41401
-31.451232
1.13555
true
0.002927
dense
-30.806589
-0.867215
-2.533527
64
0.135764
1,024
19
960
1
2.618554
0.002927
46.378033
success
0.054105
0.000001
under-trained
960
0.002927
0.002266
109
model.layers.15.self_attn.o_proj
0.073486
4,096
4,096
1
22.78947
-54.912991
1.567856
true
0.003894
dense
-54.643475
-0.702228
-2.409577
64
0.198505
4,096
14
4,032
1
5.823481
0.003894
50.974079
success
0.062404
0
under-trained
4,032
0.003894
0.003261
110
model.layers.15.self_attn.q_proj
0.038691
4,096
4,096
1
11.393952
-27.145479
1.566538
true
0.004145
dense
-27.047657
-0.808669
-2.382446
64
0.155357
4,096
27
4,032
1
2.000317
0.004145
37.478085
success
0.064384
0
under-trained
4,032
0.004145
0.002417
111
model.layers.15.self_attn.v_proj
0.09666
1,024
4,096
4
19.74294
-52.107331
1.137044
true
0.002295
dense
-51.163957
-0.894979
-2.639289
64
0.127356
1,024
26
960
1
3.675793
0.002295
55.502193
success
0.047902
0.000001
under-trained
960
0.002295
0.002011
112
model.layers.16.mlp.down_proj
0.03637
4,096
14,336
3.5
12.299618
-25.960552
1.56721
true
0.00775
dense
-25.864351
-0.513355
-2.11068
64
0.306651
4,096
27
4,032
1
2.174612
0.00775
39.566231
success
0.088036
0.000001
under-trained
4,032
0.00775
0.004679
113
model.layers.16.mlp.gate_proj
0.04882
4,096
14,336
3.5
5.786708
-11.038166
1.56046
true
0.012374
dense
-10.783695
-0.500731
-1.907503
64
0.315696
4,096
34
4,032
1
0.820914
0.012374
25.513638
success
0.111237
0.000001
4,032
0.012374
0.004401
114
model.layers.16.mlp.up_proj
0.050119
4,096
14,336
3.5
6.408145
-12.304469
1.562002
true
0.012019
dense
-12.090176
-0.489292
-1.92013
64
0.324122
4,096
32
4,032
1
0.956034
0.012019
26.9673
success
0.109631
0.000001
under-trained
4,032
0.012019
0.004628
115
model.layers.16.self_attn.k_proj
0.052562
1,024
4,096
4
7.736304
-19.403313
1.134752
true
0.003104
dense
-18.746053
-0.895812
-2.508086
64
0.127113
1,024
47
960
1
0.982591
0.003104
40.951885
success
0.055713
0.000001
under-trained
960
0.003104
0.001778
116
model.layers.16.self_attn.o_proj
0.044595
4,096
4,096
1
16.782293
-39.838156
1.567677
true
0.004228
dense
-39.667576
-0.707306
-2.373821
64
0.196198
4,096
31
4,032
1
2.834584
0.004228
46.399624
success
0.065026
0
under-trained
4,032
0.004228
0.003038
117
model.layers.16.self_attn.q_proj
0.059982
4,096
4,096
1
9.882449
-23.644621
1.566412
true
0.00405
dense
-23.432238
-0.811707
-2.392587
64
0.154274
4,096
35
4,032
1
1.501408
0.00405
38.096092
success
0.063637
0
under-trained
4,032
0.00405
0.002305
118
model.layers.16.self_attn.v_proj
0.110164
1,024
4,096
4
17.070893
-45.159406
1.137101
true
0.002263
dense
-44.087106
-0.895655
-2.645404
64
0.127158
1,024
39
960
1
2.573402
0.002263
56.201641
success
0.047566
0.000001
under-trained
960
0.002263
0.001937
119
model.layers.17.mlp.down_proj
0.045365
4,096
14,336
3.5
12.22335
-25.598136
1.566806
true
0.00805
dense
-25.536009
-0.521363
-2.0942
64
0.301049
4,096
64
4,032
1
1.402919
0.00805
37.39695
success
0.089722
0.000001
under-trained
4,032
0.00805
0.004277
120
model.layers.17.mlp.gate_proj
0.04479
4,096
14,336
3.5
6.020877
-11.465776
1.561047
true
0.012464
dense
-11.293104
-0.506241
-1.904336
64
0.311716
4,096
31
4,032
1
0.901776
0.012464
25.008944
success
0.111643
0.000001
under-trained
4,032
0.012464
0.00445
121
model.layers.17.mlp.up_proj
0.055498
4,096
14,336
3.5
6.33513
-12.266333
1.562622
true
0.011581
dense
-12.056853
-0.498333
-1.93624
64
0.317444
4,096
25
4,032
1
1.067026
0.011581
27.40988
success
0.107617
0.000001
under-trained
4,032
0.011581
0.004714
122
model.layers.17.self_attn.k_proj
0.050712
1,024
4,096
4
7.179098
-18.214098
1.134169
true
0.002903
dense
-17.593216
-0.947164
-2.537101
64
0.112937
1,024
48
960
1
0.891876
0.002903
38.898895
success
0.053883
0.000001
under-trained
960
0.002903
0.001556
123
model.layers.17.self_attn.o_proj
0.054669
4,096
4,096
1
17.646075
-43.286266
1.567636
true
0.003524
dense
-42.894703
-0.759549
-2.453025
64
0.173961
4,096
19
4,032
1
3.818872
0.003524
49.371487
success
0.059359
0
under-trained
4,032
0.003524
0.002814
124
model.layers.17.self_attn.q_proj
0.035606
4,096
4,096
1
8.667275
-20.609269
1.565403
true
0.00419
dense
-20.455829
-0.848533
-2.377826
64
0.141732
4,096
32
4,032
1
1.355396
0.00419
33.82925
success
0.064727
0
under-trained
4,032
0.00419
0.00213
125
model.layers.17.self_attn.v_proj
0.103416
1,024
4,096
4
22.441121
-59.716689
1.136952
true
0.002183
dense
-59.160964
-0.936391
-2.661039
64
0.115773
1,024
18
960
1
5.053721
0.002183
53.045414
success
0.046718
0.000001
under-trained
960
0.002183
0.001888
126
model.layers.18.mlp.down_proj
0.047888
4,096
14,336
3.5
13.341258
-28.149363
1.566923
true
0.007763
dense
-28.090671
-0.529976
-2.109948
64
0.295137
4,096
64
4,032
1
1.542657
0.007763
38.016453
success
0.08811
0.000001
under-trained
4,032
0.007763
0.004229
127
model.layers.18.mlp.gate_proj
0.047416
4,096
14,336
3.5
5.802708
-11.059557
1.559926
true
0.012419
dense
-10.841252
-0.517615
-1.905931
64
0.303658
4,096
34
4,032
1
0.823658
0.012419
24.45204
success
0.111438
0.000001
4,032
0.012419
0.00423
128
model.layers.18.mlp.up_proj
0.05985
4,096
14,336
3.5
6.70596
-12.740621
1.56102
true
0.012592
dense
-12.626483
-0.511941
-1.899895
64
0.307652
4,096
47
4,032
1
0.8323
0.012592
24.431744
success
0.112215
0.000001
under-trained
4,032
0.012592
0.004142
129
model.layers.18.self_attn.k_proj
0.108372
1,024
4,096
4
6.960606
-17.857317
1.135391
true
0.00272
dense
-16.791156
-0.89
-2.565483
64
0.128825
1,024
54
960
1
0.811136
0.00272
47.367775
success
0.05215
0.000001
under-trained
960
0.00272
0.00174
130
model.layers.18.self_attn.o_proj
0.088649
4,096
4,096
1
8.089083
-17.517363
1.561164
true
0.00683
dense
-17.390266
-0.768608
-2.165556
64
0.17037
4,096
64
4,032
1
0.886135
0.00683
24.942989
success
0.082646
0
under-trained
4,032
0.00683
0.002256
131
model.layers.18.self_attn.q_proj
0.046052
4,096
4,096
1
8.924982
-21.457456
1.566115
true
0.003943
dense
-21.25436
-0.8447
-2.404202
64
0.142988
4,096
44
4,032
1
1.194736
0.003943
36.26619
success
0.062791
0
under-trained
4,032
0.003943
0.002054
132
model.layers.18.self_attn.v_proj
0.050976
1,024
4,096
4
14.76355
-38.182098
1.136514
true
0.002593
dense
-37.95285
-0.933353
-2.586241
64
0.116586
1,024
18
960
1
3.2441
0.002593
44.966438
success
0.050919
0.000001
under-trained
960
0.002593
0.001889
133
model.layers.19.mlp.down_proj
0.052127
4,096
14,336
3.5
15.739191
-33.756519
1.567286
true
0.007166
dense
-33.731273
-0.550695
-2.144743
64
0.281388
4,096
64
4,032
1
1.842399
0.007166
39.268841
success
0.08465
0.000001
under-trained
4,032
0.007166
0.004091
134
model.layers.19.mlp.gate_proj
0.046502
4,096
14,336
3.5
5.941198
-11.533365
1.560244
true
0.011448
dense
-11.287487
-0.541057
-1.941252
64
0.287702
4,096
37
4,032
1
0.812328
0.011448
25.130192
success
0.106998
0.000001
4,032
0.011448
0.003959
135
model.layers.19.mlp.up_proj
0.058156
4,096
14,336
3.5
6.254391
-12.406597
1.561834
true
0.010383
dense
-12.083018
-0.539091
-1.983662
64
0.289008
4,096
29
4,032
1
0.975716
0.010383
27.8337
success
0.101899
0.000001
under-trained
4,032
0.010383
0.004184
136
model.layers.19.self_attn.k_proj
0.052233
1,024
4,096
4
7.138026
-18.337259
1.134431
true
0.002698
dense
-17.599834
-0.95689
-2.568954
64
0.110436
1,024
50
960
1
0.868048
0.002698
40.932098
success
0.051943
0.000001
under-trained
960
0.002698
0.001515
137
model.layers.19.self_attn.o_proj
0.075721
4,096
4,096
1
18.962458
-47.43336
1.567721
true
0.003152
dense
-47.108238
-0.806771
-2.501435
64
0.156037
4,096
16
4,032
1
4.490614
0.003152
49.506718
success
0.056141
0
under-trained
4,032
0.003152
0.002545
138
model.layers.19.self_attn.q_proj
0.049252
4,096
4,096
1
11.574794
-28.732051
1.566537
true
0.003294
dense
-28.514098
-0.873983
-2.482295
64
0.133665
4,096
22
4,032
1
2.254554
0.003294
40.579983
success
0.057392
0
under-trained
4,032
0.003294
0.002148
139
model.layers.19.self_attn.v_proj
0.112999
1,024
4,096
4
11.720165
-31.715583
1.137021
true
0.001968
dense
-30.618044
-0.971436
-2.70607
64
0.106798
1,024
63
960
1
1.350614
0.001968
54.279236
success
0.044357
0.000001
under-trained
960
0.001968
0.001519
140
model.layers.20.mlp.down_proj
0.061206
4,096
14,336
3.5
18.627453
-40.568402
1.567502
true
0.006639
dense
-40.560511
-0.577345
-2.177882
64
0.26464
4,096
63
4,032
1
2.22085
0.006639
39.860016
success
0.081481
0.000001
under-trained
4,032
0.006639
0.003898
141
model.layers.20.mlp.gate_proj
0.051517
4,096
14,336
3.5
6.108503
-12.000854
1.560681
true
0.010849
dense
-11.798363
-0.563479
-1.964615
64
0.273225
4,096
39
4,032
1
0.818015
0.010849
25.184643
success
0.104158
0.000001
under-trained
4,032
0.010849
0.003738
142
model.layers.20.mlp.up_proj
0.058065
4,096
14,336
3.5
6.712795
-13.641759
1.562814
true
0.009285
dense
-13.338223
-0.568168
-2.032203
64
0.270291
4,096
30
4,032
1
1.043009
0.009285
29.109493
success
0.09636
0.000001
under-trained
4,032
0.009285
0.003918
143
model.layers.20.self_attn.k_proj
0.041516
1,024
4,096
4
7.861621
-20.866377
1.134635
true
0.002217
dense
-20.097021
-1.026386
-2.654208
64
0.094105
1,024
37
960
1
1.128044
0.002217
42.444511
success
0.047086
0.000001
under-trained
960
0.002217
0.001376
144
model.layers.20.self_attn.o_proj
0.052326
4,096
4,096
1
15.763416
-40.379479
1.567751
true
0.002744
dense
-40.009999
-0.874711
-2.561594
64
0.133441
4,096
37
4,032
1
2.427091
0.002744
48.627666
success
0.052384
0
under-trained
4,032
0.002744
0.002028
145
model.layers.20.self_attn.q_proj
0.044572
4,096
4,096
1
8.874127
-21.844081
1.565419
true
0.003455
dense
-21.730231
-0.940223
-2.461547
64
0.114756
4,096
25
4,032
1
1.574825
0.003455
33.214188
success
0.05878
0
under-trained
4,032
0.003455
0.001787
146
model.layers.20.self_attn.v_proj
0.101759
1,024
4,096
4
23.903313
-65.707531
1.137016
true
0.001783
dense
-64.963068
-1.009508
-2.748888
64
0.097834
1,024
21
960
1
4.997913
0.001783
54.875648
success
0.042224
0.000001
under-trained
960
0.001783
0.00158
147
model.layers.21.mlp.down_proj
0.067235
4,096
14,336
3.5
20.880401
-46.272757
1.56767
true
0.00608
dense
-46.269135
-0.604594
-2.216086
64
0.248545
4,096
63
4,032
1
2.504695
0.00608
40.878166
success
0.077975
0.000001
under-trained
4,032
0.00608
0.003686
148
model.layers.21.mlp.gate_proj
0.047487
4,096
14,336
3.5
6.654224
-13.189706
1.561156
true
0.010419
dense
-13.053557
-0.584819
-1.982155
64
0.260124
4,096
47
4,032
1
0.824753
0.010419
24.965286
success
0.102076
0.000001
under-trained
4,032
0.010419
0.003499
149
model.layers.21.mlp.up_proj
0.058968
4,096
14,336
3.5
6.700749
-13.659004
1.562909
true
0.009153
dense
-13.446833
-0.590596
-2.038429
64
0.256687
4,096
28
4,032
1
1.07734
0.009153
28.043589
success
0.095672
0.000001
under-trained
4,032
0.009153
0.003754
150
model.layers.21.self_attn.k_proj
0.049508
1,024
4,096
4
6.538743
-17.295736
1.133256
true
0.002264
dense
-16.618364
-1.068307
-2.645116
64
0.085446
1,024
46
960
1
0.816643
0.002264
37.740685
success
0.047582
0.000001
under-trained
960
0.002264
0.001169
151
model.layers.21.self_attn.o_proj
0.065618
4,096
4,096
1
16.770174
-44.283508
1.567853
true
0.002288
dense
-43.920455
-0.94789
-2.640611
64
0.112748
4,096
38
4,032
1
2.55826
0.002288
49.285667
success
0.047829
0
under-trained
4,032
0.002288
0.001711
152
model.layers.21.self_attn.q_proj
0.049632
4,096
4,096
1
9.066903
-22.476881
1.564873
true
0.003319
dense
-22.418352
-0.988838
-2.479003
64
0.102603
4,096
19
4,032
1
1.850674
0.003319
30.914713
success
0.05761
0
under-trained
4,032
0.003319
0.001678
153
model.layers.21.self_attn.v_proj
0.102323
1,024
4,096
4
12.668749
-35.473683
1.136968
true
0.001585
dense
-34.520934
-1.075054
-2.800094
64
0.084129
1,024
54
960
1
1.587916
0.001585
53.093327
success
0.039806
0.000001
under-trained
960
0.001585
0.001223
154
model.layers.22.mlp.down_proj
0.065357
4,096
14,336
3.5
21.333619
-47.843349
1.567693
true
0.00572
dense
-47.83868
-0.625556
-2.242627
64
0.236834
4,096
63
4,032
1
2.561795
0.00572
41.406754
success
0.075629
0.000001
under-trained
4,032
0.00572
0.003517
155
model.layers.22.mlp.gate_proj
0.052411
4,096
14,336
3.5
6.327244
-12.577383
1.561112
true
0.010285
dense
-12.401841
-0.586662
-1.987814
64
0.259023
4,096
35
4,032
1
0.900469
0.010285
25.185551
success
0.101413
0.000001
under-trained
4,032
0.010285
0.003633
156
model.layers.22.mlp.up_proj
0.055813
4,096
14,336
3.5
6.512486
-13.175072
1.56233
true
0.009483
dense
-12.987822
-0.595345
-2.023048
64
0.253895
4,096
29
4,032
1
1.023643
0.009483
26.773323
success
0.097381
0.000001
under-trained
4,032
0.009483
0.003677
157
model.layers.22.self_attn.k_proj
0.049827
1,024
4,096
4
6.607918
-17.565098
1.132208
true
0.002197
dense
-16.957388
-1.10369
-2.65819
64
0.078761
1,024
29
960
1
1.041364
0.002197
35.850883
success
0.046871
0.000001
under-trained
960
0.002197
0.001194
158
model.layers.22.self_attn.o_proj
0.041494
4,096
4,096
1
13.129645
-34.940046
1.567293
true
0.002182
dense
-34.489469
-1.004539
-2.661157
64
0.09896
4,096
55
4,032
1
1.635561
0.002182
45.354275
success
0.046711
0
under-trained
4,032
0.002182
0.001436
159
model.layers.22.self_attn.q_proj
0.04183
4,096
4,096
1
7.694742
-19.28764
1.564005
true
0.003115
dense
-19.193126
-1.041616
-2.5066
64
0.090862
4,096
27
4,032
1
1.288404
0.003115
29.173191
success
0.055808
0
under-trained
4,032
0.003115
0.001394
160
model.layers.22.self_attn.v_proj
0.079668
1,024
4,096
4
19.708094
-55.116658
1.136799
true
0.001597
dense
-54.796141
-1.100267
-2.796651
64
0.079384
1,024
15
960
1
4.830409
0.001597
49.703156
success
0.039965
0.000001
under-trained
960
0.001597
0.001312
161
model.layers.23.mlp.down_proj
0.067607
4,096
14,336
3.5
21.848974
-49.418214
1.567716
true
0.005473
dense
-49.41545
-0.647612
-2.261809
64
0.225106
4,096
64
4,032
1
2.606122
0.005473
41.133663
success
0.073977
0.000001
under-trained
4,032
0.005473
0.003344
162
model.layers.23.mlp.gate_proj
0.056297
4,096
14,336
3.5
6.059427
-11.977099
1.560636
true
0.010553
dense
-11.811683
-0.592699
-1.976606
64
0.255447
4,096
28
4,032
1
0.956142
0.010553
24.205093
success
0.10273
0.000001
under-trained
4,032
0.010553
0.003698
163
model.layers.23.mlp.up_proj
0.063931
4,096
14,336
3.5
6.466909
-13.111112
1.561809
true
0.009388
dense
-12.900208
-0.607263
-2.027416
64
0.247023
4,096
29
4,032
1
1.015179
0.009388
26.311918
success
0.096893
0.000001
under-trained
4,032
0.009388
0.003575
164
model.layers.23.self_attn.k_proj
0.047964
1,024
4,096
4
5.956014
-16.085416
1.131418
true
0.001992
dense
-15.357696
-1.148579
-2.700702
64
0.071027
1,024
64
960
1
0.619502
0.001992
35.65517
success
0.044632
0.000001
960
0.001992
0.000886
165
model.layers.23.self_attn.o_proj
0.057257
4,096
4,096
1
9.506238
-23.880512
1.565192
true
0.003075
dense
-23.710591
-0.986447
-2.512089
64
0.10317
4,096
64
4,032
1
1.06328
0.003075
33.546062
success
0.055457
0
under-trained
4,032
0.003075
0.001416
166
model.layers.23.self_attn.q_proj
0.043697
4,096
4,096
1
6.710396
-17.197635
1.564053
true
0.002736
dense
-16.960834
-1.072318
-2.562835
64
0.084661
4,096
45
4,032
1
0.851256
0.002736
30.93972
success
0.05231
0
under-trained
4,032
0.002736
0.001165
167
model.layers.23.self_attn.v_proj
0.07277
1,024
4,096
4
12.088783
-33.232434
1.136598
true
0.001782
dense
-32.706667
-1.075935
-2.749031
64
0.083959
1,024
44
960
1
1.671697
0.001782
47.108139
success
0.042217
0.000001
under-trained
960
0.001782
0.001241
168
model.layers.24.mlp.down_proj
0.065622
4,096
14,336
3.5
21.169251
-48.697705
1.567789
true
0.005007
dense
-48.690095
-0.671119
-2.300398
64
0.213246
4,096
64
4,032
1
2.521156
0.005007
42.587193
success
0.070762
0.000001
under-trained
4,032
0.005007
0.003164
169
model.layers.24.mlp.gate_proj
0.06806
4,096
14,336
3.5
6.658305
-13.271129
1.560756
true
0.010159
dense
-13.148111
-0.613375
-1.993169
64
0.243571
4,096
39
4,032
1
0.906054
0.010159
23.976974
success
0.10079
0.000001
under-trained
4,032
0.010159
0.00337
170
model.layers.24.mlp.up_proj
0.078056
4,096
14,336
3.5
6.230828
-12.752574
1.562227
true
0.008981
dense
-12.565062
-0.631209
-2.04669
64
0.233771
4,096
22
4,032
1
1.115216
0.008981
26.030411
success
0.094767
0.000001
under-trained
4,032
0.008981
0.003496
171
model.layers.24.self_attn.k_proj
0.037613
1,024
4,096
4
5.32169
-14.107453
1.13031
true
0.002234
dense
-13.541941
-1.153247
-2.650934
64
0.070267
1,024
52
960
1
0.599311
0.002234
31.454811
success
0.047264
0.000001
960
0.002234
0.000892
172
model.layers.24.self_attn.o_proj
0.044291
4,096
4,096
1
13.472331
-36.547637
1.56741
true
0.001937
dense
-36.173019
-1.054174
-2.712792
64
0.088273
4,096
58
4,032
1
1.637697
0.001937
45.563667
success
0.044015
0
under-trained
4,032
0.001937
0.001278
173
model.layers.24.self_attn.q_proj
0.039975
4,096
4,096
1
7.641116
-19.853336
1.564035
true
0.002522
dense
-19.672138
-1.105511
-2.598225
64
0.078431
4,096
30
4,032
1
1.212496
0.002522
31.096643
success
0.050221
0
under-trained
4,032
0.002522
0.001185
174
model.layers.24.self_attn.v_proj
0.087479
1,024
4,096
4
22.192286
-62.33312
1.136846
true
0.001553
dense
-61.984562
-1.103115
-2.808774
64
0.078865
1,024
13
960
1
5.877683
0.001553
50.776108
success
0.039411
0.000001
under-trained
960
0.001553
0.001317
175
model.layers.25.mlp.down_proj
0.048318
4,096
14,336
3.5
20.719397
-47.925841
1.567751
true
0.004863
dense
-47.917582
-0.686748
-2.313091
64
0.205709
4,096
64
4,032
1
2.464925
0.004863
42.300262
success
0.069736
0.000001
under-trained
4,032
0.004863
0.003048
176
model.layers.25.mlp.gate_proj
0.069656
4,096
14,336
3.5
6.934986
-13.859512
1.561265
true
0.010035
dense
-13.775954
-0.623837
-1.998492
64
0.237773
4,096
35
4,032
1
1.003196
0.010035
23.694872
success
0.100174
0.000001
under-trained
4,032
0.010035
0.003364
177
model.layers.25.mlp.up_proj
0.099564
4,096
14,336
3.5
7.081751
-14.561236
1.562385
true
0.008787
dense
-14.43802
-0.646459
-2.056163
64
0.225705
4,096
27
4,032
1
1.170433
0.008787
25.686459
success
0.093739
0.000001
under-trained
4,032
0.008787
0.003317
178
model.layers.25.self_attn.k_proj
0.055758
1,024
4,096
4
7.444985
-20.07066
1.133036
true
0.002014
dense
-19.664577
-1.152309
-2.695863
64
0.070419
1,024
25
960
1
1.288997
0.002014
34.958603
success
0.044882
0.000001
under-trained
960
0.002014
0.00111
179
model.layers.25.self_attn.o_proj
0.050357
4,096
4,096
1
11.625005
-31.168731
1.566863
true
0.002084
dense
-30.821281
-1.057331
-2.68118
64
0.087633
4,096
64
4,032
1
1.328126
0.002084
42.058033
success
0.045647
0
under-trained
4,032
0.002084
0.001239
180
model.layers.25.self_attn.q_proj
0.049175
4,096
4,096
1
6.111549
-16.078789
1.564285
true
0.002339
dense
-15.621483
-1.100344
-2.630886
64
0.07937
4,096
64
4,032
1
0.638944
0.002339
33.926746
success
0.048368
0
under-trained
4,032
0.002339
0.001004
181
model.layers.25.self_attn.v_proj
0.061502
1,024
4,096
4
15.363986
-43.231103
1.13677
true
0.001535
dense
-42.63318
-1.112548
-2.813795
64
0.077171
1,024
28
960
1
2.714538
0.001535
50.262863
success
0.039183
0.000001
under-trained
960
0.001535
0.001205
182
model.layers.26.mlp.down_proj
0.055896
4,096
14,336
3.5
19.291067
-45.162404
1.567739
true
0.004559
dense
-45.122915
-0.700354
-2.341105
64
0.199364
4,096
62
4,032
1
2.322968
0.004559
43.727108
success
0.067522
0.000001
under-trained
4,032
0.004559
0.002947
183
model.layers.26.mlp.gate_proj
0.085826
4,096
14,336
3.5
6.934983
-13.885229
1.561158
true
0.009949
dense
-13.801381
-0.631356
-2.002201
64
0.233692
4,096
33
4,032
1
1.033148
0.009949
23.487938
success
0.099747
0.000001
under-trained
4,032
0.009949
0.003328
184
model.layers.26.mlp.up_proj
0.104035
4,096
14,336
3.5
6.006644
-12.376087
1.562014
true
0.008702
dense
-12.173414
-0.654805
-2.0604
64
0.221409
4,096
18
4,032
1
1.180077
0.008702
25.444557
success
0.093283
0.000001
under-trained
4,032
0.008702
0.003379
185
model.layers.26.self_attn.k_proj
0.056932
1,024
4,096
4
5.374071
-14.446276
1.131228
true
0.00205
dense
-13.682997
-1.136951
-2.688144
64
0.072954
1,024
50
960
1
0.618587
0.00205
35.578926
success
0.045282
0.000001
960
0.00205
0.00094
186
model.layers.26.self_attn.o_proj
0.088726
4,096
4,096
1
4.214271
-9.801506
1.554751
true
0.004723
dense
-9.488855
-1.02519
-2.325789
64
0.094365
4,096
17
4,032
1
0.779575
0.004723
19.980146
success
0.068724
0
4,032
0.004723
0.001383
187
model.layers.26.self_attn.q_proj
0.043501
4,096
4,096
1
6.091549
-15.42408
1.562849
true
0.002937
dense
-15.203184
-1.081572
-2.532046
64
0.082876
4,096
52
4,032
1
0.706071
0.002937
28.214588
success
0.054197
0
under-trained
4,032
0.002937
0.001088
188
model.layers.26.self_attn.v_proj
0.054334
1,024
4,096
4
13.788339
-38.606445
1.136559
true
0.001585
dense
-38.187704
-1.129555
-2.799934
64
0.074207
1,024
26
960
1
2.508
0.001585
46.814354
success
0.039814
0.000001
under-trained
960
0.001585
0.001162
189
model.layers.27.mlp.down_proj
0.06611
4,096
14,336
3.5
17.607994
-41.717967
1.567614
true
0.004273
dense
-41.524366
-0.71679
-2.369263
64
0.19196
4,096
64
4,032
1
2.075999
0.004273
44.92347
success
0.065369
0.000001
under-trained
4,032
0.004273
0.002816
190
model.layers.27.mlp.gate_proj
0.085341
4,096
14,336
3.5
7.507809
-14.940597
1.560531
true
0.010233
dense
-14.876542
-0.630876
-1.990008
64
0.23395
4,096
51
4,032
1
0.911276
0.010233
22.862886
success
0.101157
0.000001
under-trained
4,032
0.010233
0.003158
191
model.layers.27.mlp.up_proj
0.090429
4,096
14,336
3.5
6.337803
-13.036704
1.561669
true
0.008771
dense
-12.855776
-0.657381
-2.056975
64
0.220099
4,096
23
4,032
1
1.113009
0.008771
25.09539
success
0.093651
0.000001
under-trained
4,032
0.008771
0.003267
192
model.layers.27.self_attn.k_proj
0.062667
1,024
4,096
4
6.433864
-17.394814
1.131266
true
0.001979
dense
-16.810661
-1.179399
-2.703634
64
0.066161
1,024
64
960
1
0.679233
0.001979
33.43763
success
0.044482
0.000001
under-trained
960
0.001979
0.00084
193
model.layers.27.self_attn.o_proj
0.052241
4,096
4,096
1
7.273309
-18.565491
1.5641
true
0.002802
dense
-18.431408
-1.090858
-2.552551
64
0.081123
4,096
28
4,032
1
1.185544
0.002802
28.952957
success
0.052933
0
under-trained
4,032
0.002802
0.001198
194
model.layers.27.self_attn.q_proj
0.042504
4,096
4,096
1
5.579411
-14.485803
1.562452
true
0.002533
dense
-14.109143
-1.120044
-2.596296
64
0.07585
4,096
64
4,032
1
0.572426
0.002533
29.940044
success
0.050333
0
4,032
0.002533
0.000931
195
model.layers.27.self_attn.v_proj
0.047939
1,024
4,096
4
12.779446
-35.765415
1.136394
true
0.00159
dense
-35.399751
-1.146404
-2.798667
64
0.071383
1,024
29
960
1
2.187388
0.00159
44.901775
success
0.039872
0.000001
under-trained
960
0.00159
0.001103
196
model.layers.28.mlp.down_proj
0.04537
4,096
14,336
3.5
15.507892
-36.617226
1.567459
true
0.004353
dense
-36.349747
-0.712012
-2.361199
64
0.194083
4,096
64
4,032
1
1.813487
0.004353
44.584877
success
0.065978
0.000001
under-trained
4,032
0.004353
0.00282
197
model.layers.28.mlp.gate_proj
0.055398
4,096
14,336
3.5
7.201658
-14.241886
1.559661
true
0.01053
dense
-14.181808
-0.636204
-1.977584
64
0.231098
4,096
63
4,032
1
0.781335
0.01053
21.947275
success
0.102614
0.000001
under-trained
4,032
0.01053
0.002993
198
model.layers.28.mlp.up_proj
0.070715
4,096
14,336
3.5
6.613636
-13.41884
1.561279
true
0.009355
dense
-13.332549
-0.659056
-2.028966
64
0.219252
4,096
30
4,032
1
1.024905
0.009355
23.437424
success
0.09672
0.000001
under-trained
4,032
0.009355
0.003147
199
model.layers.28.self_attn.k_proj
0.037904
1,024
4,096
4
5.71132
-15.440612
1.131899
true
0.001979
dense
-14.690031
-1.143304
-2.70351
64
0.071895
1,024
63
960
1
0.593571
0.001979
36.325069
success
0.044488
0.000001
960
0.001979
0.000892