layer_id
int64 0
223
| name
stringlengths 26
32
| D
float64 0.02
0.18
| M
int64 1.02k
4.1k
| N
int64 4.1k
14.3k
| Q
float64 1
4
| alpha
float64 3.57
40.4
| alpha_weighted
float64 -123.19
-7.38
| entropy
float64 1.1
1.57
| has_esd
bool 1
class | lambda_max
float32 0
0.01
| layer_type
stringclasses 1
value | log_alpha_norm
float64 -123.03
-6.97
| log_norm
float32 -1.93
-0.76
| log_spectral_norm
float32 -3.35
-2
| matrix_rank
int64 64
64
| norm
float32 0.01
0.17
| num_evals
int64 1.02k
4.1k
| num_pl_spikes
int64 7
64
| rank_loss
int64 960
4.03k
| rf
int64 1
1
| sigma
float64 0.59
11.9
| spectral_norm
float32 0
0.01
| stable_rank
float32 14
56.5
| status
stringclasses 1
value | sv_max
float64 0.02
0.1
| sv_min
float64 0
0
| warning
stringclasses 2
values | weak_rank_loss
int64 960
4.03k
| xmax
float64 0
0.01
| xmin
float64 0
0
|
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
100
|
model.layers.14.mlp.up_proj
| 0.070595
| 4,096
| 14,336
| 3.5
| 5.295959
| -11.124345
| 1.560849
| true
| 0.007934
|
dense
| -11.020632
| -0.761963
| -2.100535
| 64
| 0.172996
| 4,096
| 17
| 4,032
| 1
| 1.041923
| 0.007934
| 21.805763
|
success
| 0.08907
| 0.000001
| 4,032
| 0.007934
| 0.002631
|
|
101
|
model.layers.14.self_attn.k_proj
| 0.108818
| 1,024
| 4,096
| 4
| 9.374769
| -27.164726
| 1.136578
| true
| 0.001266
|
dense
| -26.117086
| -1.18827
| -2.897642
| 64
| 0.064823
| 1,024
| 60
| 960
| 1
| 1.081178
| 0.001266
| 51.21207
|
success
| 0.035578
| 0.000001
|
under-trained
| 960
| 0.001266
| 0.000903
|
102
|
model.layers.14.self_attn.o_proj
| 0.063818
| 4,096
| 4,096
| 1
| 10.309597
| -27.056012
| 1.565609
| true
| 0.002375
|
dense
| -26.82605
| -1.075338
| -2.624352
| 64
| 0.084074
| 4,096
| 64
| 4,032
| 1
| 1.1637
| 0.002375
| 35.400898
|
success
| 0.048733
| 0
|
under-trained
| 4,032
| 0.002375
| 0.001168
|
103
|
model.layers.14.self_attn.q_proj
| 0.038597
| 4,096
| 4,096
| 1
| 10.648016
| -27.996992
| 1.566439
| true
| 0.002348
|
dense
| -27.951509
| -1.095221
| -2.629315
| 64
| 0.080312
| 4,096
| 64
| 4,032
| 1
| 1.206002
| 0.002348
| 34.20536
|
success
| 0.048455
| 0
|
under-trained
| 4,032
| 0.002348
| 0.001123
|
104
|
model.layers.14.self_attn.v_proj
| 0.063967
| 1,024
| 4,096
| 4
| 18.977392
| -56.967035
| 1.136897
| true
| 0.000996
|
dense
| -56.612306
| -1.303714
| -3.001837
| 64
| 0.049692
| 1,024
| 21
| 960
| 1
| 3.922989
| 0.000996
| 49.902519
|
success
| 0.031556
| 0.000001
|
under-trained
| 960
| 0.000996
| 0.000796
|
105
|
model.layers.15.mlp.down_proj
| 0.030025
| 4,096
| 14,336
| 3.5
| 19.888242
| -49.551639
| 1.567968
| true
| 0.003225
|
dense
| -49.298357
| -0.794197
| -2.491504
| 64
| 0.160621
| 4,096
| 64
| 4,032
| 1
| 2.36103
| 0.003225
| 49.808956
|
success
| 0.056787
| 0.000001
|
under-trained
| 4,032
| 0.003225
| 0.002376
|
106
|
model.layers.15.mlp.gate_proj
| 0.068083
| 4,096
| 14,336
| 3.5
| 5.924201
| -12.518264
| 1.561087
| true
| 0.007708
|
dense
| -12.462627
| -0.777118
| -2.113072
| 64
| 0.167064
| 4,096
| 22
| 4,032
| 1
| 1.049843
| 0.007708
| 21.674715
|
success
| 0.087794
| 0.000001
| 4,032
| 0.007708
| 0.002473
|
|
107
|
model.layers.15.mlp.up_proj
| 0.079643
| 4,096
| 14,336
| 3.5
| 5.669253
| -12.220102
| 1.562827
| true
| 0.00699
|
dense
| -12.101856
| -0.76957
| -2.155505
| 64
| 0.169992
| 4,096
| 12
| 4,032
| 1
| 1.347897
| 0.00699
| 24.318357
|
success
| 0.083608
| 0.000001
| 4,032
| 0.00699
| 0.002777
|
|
108
|
model.layers.15.self_attn.k_proj
| 0.101475
| 1,024
| 4,096
| 4
| 10.378981
| -29.608627
| 1.136357
| true
| 0.001404
|
dense
| -28.722908
| -1.157424
| -2.852749
| 64
| 0.069595
| 1,024
| 42
| 960
| 1
| 1.447208
| 0.001404
| 49.582108
|
success
| 0.037465
| 0.000001
|
under-trained
| 960
| 0.001404
| 0.001026
|
109
|
model.layers.15.self_attn.o_proj
| 0.058208
| 4,096
| 4,096
| 1
| 22.957791
| -64.240785
| 1.567953
| true
| 0.001591
|
dense
| -64.025073
| -1.092063
| -2.798213
| 64
| 0.080898
| 4,096
| 18
| 4,032
| 1
| 5.175501
| 0.001591
| 50.83342
|
success
| 0.039893
| 0
|
under-trained
| 4,032
| 0.001591
| 0.0013
|
110
|
model.layers.15.self_attn.q_proj
| 0.058694
| 4,096
| 4,096
| 1
| 13.271464
| -35.264829
| 1.567114
| true
| 0.002202
|
dense
| -35.242807
| -1.089259
| -2.657192
| 64
| 0.081422
| 4,096
| 46
| 4,032
| 1
| 1.809329
| 0.002202
| 36.977089
|
success
| 0.046925
| 0
|
under-trained
| 4,032
| 0.002202
| 0.001202
|
111
|
model.layers.15.self_attn.v_proj
| 0.11419
| 1,024
| 4,096
| 4
| 22.354517
| -68.653411
| 1.137139
| true
| 0.000849
|
dense
| -67.828455
| -1.325262
| -3.07112
| 64
| 0.047287
| 1,024
| 26
| 960
| 1
| 4.187965
| 0.000849
| 55.700417
|
success
| 0.029137
| 0.000001
|
under-trained
| 960
| 0.000849
| 0.000747
|
112
|
model.layers.16.mlp.down_proj
| 0.064003
| 4,096
| 14,336
| 3.5
| 18.423216
| -44.586561
| 1.567581
| true
| 0.003801
|
dense
| -44.537154
| -0.789763
| -2.420129
| 64
| 0.162269
| 4,096
| 64
| 4,032
| 1
| 2.177902
| 0.003801
| 42.693913
|
success
| 0.06165
| 0.000001
|
under-trained
| 4,032
| 0.003801
| 0.002387
|
113
|
model.layers.16.mlp.gate_proj
| 0.095854
| 4,096
| 14,336
| 3.5
| 6.133836
| -12.7012
| 1.56011
| true
| 0.008498
|
dense
| -12.670542
| -0.765247
| -2.070678
| 64
| 0.171693
| 4,096
| 25
| 4,032
| 1
| 1.026767
| 0.008498
| 20.203686
|
success
| 0.092185
| 0.000001
|
under-trained
| 4,032
| 0.008498
| 0.002498
|
114
|
model.layers.16.mlp.up_proj
| 0.09054
| 4,096
| 14,336
| 3.5
| 6.53151
| -13.916047
| 1.562146
| true
| 0.007403
|
dense
| -13.871134
| -0.768526
| -2.130602
| 64
| 0.170402
| 4,096
| 24
| 4,032
| 1
| 1.129115
| 0.007403
| 23.018433
|
success
| 0.08604
| 0.000001
|
under-trained
| 4,032
| 0.007403
| 0.002511
|
115
|
model.layers.16.self_attn.k_proj
| 0.115521
| 1,024
| 4,096
| 4
| 9.912147
| -28.712586
| 1.136763
| true
| 0.001269
|
dense
| -27.50179
| -1.163768
| -2.896707
| 64
| 0.068585
| 1,024
| 60
| 960
| 1
| 1.150553
| 0.001269
| 54.067829
|
success
| 0.035616
| 0.000001
|
under-trained
| 960
| 0.001269
| 0.000962
|
116
|
model.layers.16.self_attn.o_proj
| 0.054619
| 4,096
| 4,096
| 1
| 21.033314
| -58.536135
| 1.567931
| true
| 0.001648
|
dense
| -58.320719
| -1.082587
| -2.78302
| 64
| 0.082682
| 4,096
| 26
| 4,032
| 1
| 3.928856
| 0.001648
| 50.168678
|
success
| 0.040597
| 0
|
under-trained
| 4,032
| 0.001648
| 0.001299
|
117
|
model.layers.16.self_attn.q_proj
| 0.045936
| 4,096
| 4,096
| 1
| 13.197938
| -34.586
| 1.56682
| true
| 0.002396
|
dense
| -34.576094
| -1.080902
| -2.620561
| 64
| 0.083004
| 4,096
| 41
| 4,032
| 1
| 1.904998
| 0.002396
| 34.646488
|
success
| 0.048946
| 0
|
under-trained
| 4,032
| 0.002396
| 0.001237
|
118
|
model.layers.16.self_attn.v_proj
| 0.078067
| 1,024
| 4,096
| 4
| 40.403046
| -123.190286
| 1.137152
| true
| 0.000893
|
dense
| -123.031911
| -1.316324
| -3.049035
| 64
| 0.04827
| 1,024
| 11
| 960
| 1
| 11.880465
| 0.000893
| 54.039471
|
success
| 0.029887
| 0.000001
|
under-trained
| 960
| 0.000893
| 0.000805
|
119
|
model.layers.17.mlp.down_proj
| 0.06046
| 4,096
| 14,336
| 3.5
| 17.860272
| -43.56073
| 1.567649
| true
| 0.003639
|
dense
| -43.471237
| -0.794595
| -2.438973
| 64
| 0.160474
| 4,096
| 64
| 4,032
| 1
| 2.107534
| 0.003639
| 44.093876
|
success
| 0.060327
| 0.000001
|
under-trained
| 4,032
| 0.003639
| 0.002356
|
120
|
model.layers.17.mlp.gate_proj
| 0.082128
| 4,096
| 14,336
| 3.5
| 5.495079
| -11.656814
| 1.561849
| true
| 0.007563
|
dense
| -11.56735
| -0.772201
| -2.121319
| 64
| 0.168966
| 4,096
| 13
| 4,032
| 1
| 1.246711
| 0.007563
| 22.341782
|
success
| 0.086964
| 0.000001
| 4,032
| 0.007563
| 0.0027
|
|
121
|
model.layers.17.mlp.up_proj
| 0.063486
| 4,096
| 14,336
| 3.5
| 6.23721
| -13.59188
| 1.563557
| true
| 0.00662
|
dense
| -13.499983
| -0.775867
| -2.17916
| 64
| 0.167545
| 4,096
| 14
| 4,032
| 1
| 1.399703
| 0.00662
| 25.310049
|
success
| 0.081362
| 0.000001
|
under-trained
| 4,032
| 0.00662
| 0.002663
|
122
|
model.layers.17.self_attn.k_proj
| 0.05194
| 1,024
| 4,096
| 4
| 13.234445
| -38.008686
| 1.136307
| true
| 0.001343
|
dense
| -37.564366
| -1.20944
| -2.871952
| 64
| 0.061739
| 1,024
| 21
| 960
| 1
| 2.669775
| 0.001343
| 45.973885
|
success
| 0.036646
| 0.000001
|
under-trained
| 960
| 0.001343
| 0.000991
|
123
|
model.layers.17.self_attn.o_proj
| 0.033461
| 4,096
| 4,096
| 1
| 14.530256
| -40.348406
| 1.567569
| true
| 0.001672
|
dense
| -39.98802
| -1.106861
| -2.776855
| 64
| 0.078188
| 4,096
| 37
| 4,032
| 1
| 2.22436
| 0.001672
| 46.772816
|
success
| 0.040886
| 0
|
under-trained
| 4,032
| 0.001672
| 0.001184
|
124
|
model.layers.17.self_attn.q_proj
| 0.034492
| 4,096
| 4,096
| 1
| 11.908328
| -31.226268
| 1.566418
| true
| 0.002387
|
dense
| -31.21486
| -1.107816
| -2.622221
| 64
| 0.078016
| 4,096
| 47
| 4,032
| 1
| 1.591143
| 0.002387
| 32.689209
|
success
| 0.048853
| 0
|
under-trained
| 4,032
| 0.002387
| 0.001138
|
125
|
model.layers.17.self_attn.v_proj
| 0.090111
| 1,024
| 4,096
| 4
| 25.721496
| -78.962157
| 1.137121
| true
| 0.000851
|
dense
| -78.326813
| -1.329413
| -3.06989
| 64
| 0.046837
| 1,024
| 21
| 960
| 1
| 5.394673
| 0.000851
| 55.01442
|
success
| 0.029178
| 0.000001
|
under-trained
| 960
| 0.000851
| 0.000753
|
126
|
model.layers.18.mlp.down_proj
| 0.035721
| 4,096
| 14,336
| 3.5
| 20.031384
| -49.405232
| 1.567859
| true
| 0.003417
|
dense
| -49.336911
| -0.806355
| -2.466391
| 64
| 0.156187
| 4,096
| 64
| 4,032
| 1
| 2.378923
| 0.003417
| 45.712662
|
success
| 0.058453
| 0.000001
|
under-trained
| 4,032
| 0.003417
| 0.002311
|
127
|
model.layers.18.mlp.gate_proj
| 0.063541
| 4,096
| 14,336
| 3.5
| 5.057453
| -10.494784
| 1.559962
| true
| 0.008412
|
dense
| -10.422957
| -0.782374
| -2.075113
| 64
| 0.165054
| 4,096
| 14
| 4,032
| 1
| 1.0844
| 0.008412
| 19.621803
|
success
| 0.091716
| 0.000001
| 4,032
| 0.008412
| 0.00255
|
|
128
|
model.layers.18.mlp.up_proj
| 0.063435
| 4,096
| 14,336
| 3.5
| 5.346702
| -11.310891
| 1.561263
| true
| 0.007665
|
dense
| -11.227929
| -0.787416
| -2.115489
| 64
| 0.163149
| 4,096
| 14
| 4,032
| 1
| 1.161705
| 0.007665
| 21.284979
|
success
| 0.08755
| 0.000001
| 4,032
| 0.007665
| 0.002524
|
|
129
|
model.layers.18.self_attn.k_proj
| 0.113236
| 1,024
| 4,096
| 4
| 7.800461
| -22.423279
| 1.136252
| true
| 0.001335
|
dense
| -21.267073
| -1.168217
| -2.874609
| 64
| 0.067887
| 1,024
| 62
| 960
| 1
| 0.863659
| 0.001335
| 50.861946
|
success
| 0.036534
| 0.000001
|
under-trained
| 960
| 0.001335
| 0.000915
|
130
|
model.layers.18.self_attn.o_proj
| 0.073369
| 4,096
| 4,096
| 1
| 11.414136
| -29.839959
| 1.565693
| true
| 0.002431
|
dense
| -29.820689
| -1.120096
| -2.614299
| 64
| 0.075841
| 4,096
| 64
| 4,032
| 1
| 1.301767
| 0.002431
| 31.203423
|
success
| 0.0493
| 0
|
under-trained
| 4,032
| 0.002431
| 0.001066
|
131
|
model.layers.18.self_attn.q_proj
| 0.053146
| 4,096
| 4,096
| 1
| 12.021277
| -31.476889
| 1.566602
| true
| 0.002408
|
dense
| -31.460674
| -1.089056
| -2.618431
| 64
| 0.08146
| 4,096
| 48
| 4,032
| 1
| 1.590784
| 0.002408
| 33.835728
|
success
| 0.049066
| 0
|
under-trained
| 4,032
| 0.002408
| 0.001188
|
132
|
model.layers.18.self_attn.v_proj
| 0.074868
| 1,024
| 4,096
| 4
| 31.943612
| -97.858903
| 1.137107
| true
| 0.000864
|
dense
| -97.507446
| -1.330271
| -3.063489
| 64
| 0.046744
| 1,024
| 12
| 960
| 1
| 8.932651
| 0.000864
| 54.102608
|
success
| 0.029394
| 0.000001
|
under-trained
| 960
| 0.000864
| 0.000776
|
133
|
model.layers.19.mlp.down_proj
| 0.052651
| 4,096
| 14,336
| 3.5
| 23.80037
| -59.519345
| 1.568035
| true
| 0.003157
|
dense
| -59.475479
| -0.821488
| -2.500774
| 64
| 0.150839
| 4,096
| 63
| 4,032
| 1
| 2.872577
| 0.003157
| 47.78442
|
success
| 0.056184
| 0.000001
|
under-trained
| 4,032
| 0.003157
| 0.002254
|
134
|
model.layers.19.mlp.gate_proj
| 0.074004
| 4,096
| 14,336
| 3.5
| 5.484326
| -11.727247
| 1.561759
| true
| 0.007272
|
dense
| -11.654583
| -0.803156
| -2.13832
| 64
| 0.157342
| 4,096
| 14
| 4,032
| 1
| 1.198487
| 0.007272
| 21.635359
|
success
| 0.085279
| 0.000001
| 4,032
| 0.007272
| 0.002451
|
|
135
|
model.layers.19.mlp.up_proj
| 0.081407
| 4,096
| 14,336
| 3.5
| 5.57579
| -12.189419
| 1.563044
| true
| 0.006514
|
dense
| -12.085657
| -0.812214
| -2.186133
| 64
| 0.154094
| 4,096
| 11
| 4,032
| 1
| 1.379653
| 0.006514
| 23.654802
|
success
| 0.080711
| 0.000001
| 4,032
| 0.006514
| 0.002491
|
|
136
|
model.layers.19.self_attn.k_proj
| 0.103896
| 1,024
| 4,096
| 4
| 9.422241
| -27.63405
| 1.136375
| true
| 0.001167
|
dense
| -26.58249
| -1.22558
| -2.932853
| 64
| 0.059487
| 1,024
| 50
| 960
| 1
| 1.191085
| 0.001167
| 50.965183
|
success
| 0.034164
| 0.000001
|
under-trained
| 960
| 0.001167
| 0.000849
|
137
|
model.layers.19.self_attn.o_proj
| 0.044688
| 4,096
| 4,096
| 1
| 21.054933
| -59.327508
| 1.567926
| true
| 0.001521
|
dense
| -59.104358
| -1.11806
| -2.817749
| 64
| 0.076197
| 4,096
| 23
| 4,032
| 1
| 4.181743
| 0.001521
| 50.082825
|
success
| 0.039005
| 0
|
under-trained
| 4,032
| 0.001521
| 0.001205
|
138
|
model.layers.19.self_attn.q_proj
| 0.043086
| 4,096
| 4,096
| 1
| 13.007356
| -34.852457
| 1.566701
| true
| 0.002092
|
dense
| -34.840439
| -1.140514
| -2.679442
| 64
| 0.072358
| 4,096
| 32
| 4,032
| 1
| 2.122621
| 0.002092
| 34.588226
|
success
| 0.045738
| 0
|
under-trained
| 4,032
| 0.002092
| 0.001104
|
139
|
model.layers.19.self_attn.v_proj
| 0.085536
| 1,024
| 4,096
| 4
| 32.112979
| -97.924641
| 1.137149
| true
| 0.000893
|
dense
| -97.656135
| -1.317667
| -3.049379
| 64
| 0.048121
| 1,024
| 17
| 960
| 1
| 7.546006
| 0.000893
| 53.915321
|
success
| 0.029875
| 0.000001
|
under-trained
| 960
| 0.000893
| 0.000785
|
140
|
model.layers.20.mlp.down_proj
| 0.06154
| 4,096
| 14,336
| 3.5
| 25.728871
| -64.630048
| 1.568064
| true
| 0.003076
|
dense
| -64.612966
| -0.836834
| -2.511966
| 64
| 0.145602
| 4,096
| 63
| 4,032
| 1
| 3.115545
| 0.003076
| 47.329498
|
success
| 0.055465
| 0.000001
|
under-trained
| 4,032
| 0.003076
| 0.002184
|
141
|
model.layers.20.mlp.gate_proj
| 0.07428
| 4,096
| 14,336
| 3.5
| 5.7562
| -12.622673
| 1.56318
| true
| 0.006414
|
dense
| -12.538929
| -0.821344
| -2.192883
| 64
| 0.150888
| 4,096
| 12
| 4,032
| 1
| 1.372997
| 0.006414
| 23.525513
|
success
| 0.080086
| 0.000001
| 4,032
| 0.006414
| 0.002385
|
|
142
|
model.layers.20.mlp.up_proj
| 0.067795
| 4,096
| 14,336
| 3.5
| 6.416853
| -14.515735
| 1.564549
| true
| 0.005469
|
dense
| -14.415343
| -0.835165
| -2.262127
| 64
| 0.146162
| 4,096
| 13
| 4,032
| 1
| 1.502365
| 0.005469
| 26.727709
|
success
| 0.07395
| 0.000001
|
under-trained
| 4,032
| 0.005469
| 0.002298
|
143
|
model.layers.20.self_attn.k_proj
| 0.084596
| 1,024
| 4,096
| 4
| 16.050775
| -47.779777
| 1.136508
| true
| 0.001055
|
dense
| -47.116229
| -1.274841
| -2.976789
| 64
| 0.053108
| 1,024
| 16
| 960
| 1
| 3.762694
| 0.001055
| 50.344112
|
success
| 0.032479
| 0.000001
|
under-trained
| 960
| 0.001055
| 0.00088
|
144
|
model.layers.20.self_attn.o_proj
| 0.095816
| 4,096
| 4,096
| 1
| 35.030403
| -101.934399
| 1.568151
| true
| 0.001231
|
dense
| -101.410505
| -1.157785
| -2.909884
| 64
| 0.069537
| 4,096
| 16
| 4,032
| 1
| 8.507601
| 0.001231
| 56.506565
|
success
| 0.03508
| 0
|
under-trained
| 4,032
| 0.001231
| 0.001121
|
145
|
model.layers.20.self_attn.q_proj
| 0.043583
| 4,096
| 4,096
| 1
| 11.930345
| -32.120584
| 1.566411
| true
| 0.002031
|
dense
| -32.11039
| -1.180249
| -2.692343
| 64
| 0.066032
| 4,096
| 39
| 4,032
| 1
| 1.750256
| 0.002031
| 32.515812
|
success
| 0.045064
| 0
|
under-trained
| 4,032
| 0.002031
| 0.000981
|
146
|
model.layers.20.self_attn.v_proj
| 0.113966
| 1,024
| 4,096
| 4
| 14.136786
| -43.462944
| 1.137104
| true
| 0.000842
|
dense
| -42.349018
| -1.328369
| -3.074457
| 64
| 0.04695
| 1,024
| 53
| 960
| 1
| 1.804476
| 0.000842
| 55.729904
|
success
| 0.029025
| 0.000001
|
under-trained
| 960
| 0.000842
| 0.00069
|
147
|
model.layers.21.mlp.down_proj
| 0.069908
| 4,096
| 14,336
| 3.5
| 25.507698
| -64.295016
| 1.568023
| true
| 0.003016
|
dense
| -64.282851
| -0.853733
| -2.520612
| 64
| 0.140045
| 4,096
| 61
| 4,032
| 1
| 3.137889
| 0.003016
| 46.438583
|
success
| 0.054915
| 0.000001
|
under-trained
| 4,032
| 0.003016
| 0.002102
|
148
|
model.layers.21.mlp.gate_proj
| 0.108281
| 4,096
| 14,336
| 3.5
| 5.774525
| -12.953684
| 1.564314
| true
| 0.005712
|
dense
| -12.833288
| -0.836879
| -2.243247
| 64
| 0.145587
| 4,096
| 8
| 4,032
| 1
| 1.688049
| 0.005712
| 25.489885
|
success
| 0.075575
| 0.000001
| 4,032
| 0.005712
| 0.002424
|
|
149
|
model.layers.21.mlp.up_proj
| 0.093581
| 4,096
| 14,336
| 3.5
| 6.776335
| -15.681946
| 1.565386
| true
| 0.00485
|
dense
| -15.571955
| -0.857898
| -2.314222
| 64
| 0.138708
| 4,096
| 11
| 4,032
| 1
| 1.741631
| 0.00485
| 28.597256
|
success
| 0.069645
| 0.000001
|
under-trained
| 4,032
| 0.00485
| 0.002209
|
150
|
model.layers.21.self_attn.k_proj
| 0.059948
| 1,024
| 4,096
| 4
| 11.910162
| -35.071107
| 1.136015
| true
| 0.001136
|
dense
| -34.693005
| -1.302377
| -2.944637
| 64
| 0.049845
| 1,024
| 22
| 960
| 1
| 2.326054
| 0.001136
| 43.879395
|
success
| 0.033704
| 0.000001
|
under-trained
| 960
| 0.001136
| 0.000798
|
151
|
model.layers.21.self_attn.o_proj
| 0.074732
| 4,096
| 4,096
| 1
| 36.651508
| -107.643644
| 1.568171
| true
| 0.001156
|
dense
| -107.320781
| -1.191226
| -2.93695
| 64
| 0.064383
| 4,096
| 13
| 4,032
| 1
| 9.887949
| 0.001156
| 55.683193
|
success
| 0.034004
| 0
|
under-trained
| 4,032
| 0.001156
| 0.001042
|
152
|
model.layers.21.self_attn.q_proj
| 0.047363
| 4,096
| 4,096
| 1
| 11.346615
| -30.906445
| 1.566245
| true
| 0.001889
|
dense
| -30.892907
| -1.217366
| -2.723847
| 64
| 0.060623
| 4,096
| 46
| 4,032
| 1
| 1.525525
| 0.001889
| 32.098225
|
success
| 0.043459
| 0
|
under-trained
| 4,032
| 0.001889
| 0.000883
|
153
|
model.layers.21.self_attn.v_proj
| 0.113763
| 1,024
| 4,096
| 4
| 14.337243
| -44.625051
| 1.137172
| true
| 0.000772
|
dense
| -43.490099
| -1.362799
| -3.112527
| 64
| 0.043371
| 1,024
| 56
| 960
| 1
| 1.782264
| 0.000772
| 56.198883
|
success
| 0.02778
| 0.000001
|
under-trained
| 960
| 0.000772
| 0.000635
|
154
|
model.layers.22.mlp.down_proj
| 0.066965
| 4,096
| 14,336
| 3.5
| 26.614543
| -68.025058
| 1.568099
| true
| 0.00278
|
dense
| -67.988984
| -0.866974
| -2.555936
| 64
| 0.13584
| 4,096
| 62
| 4,032
| 1
| 3.25305
| 0.00278
| 48.86095
|
success
| 0.052727
| 0.000001
|
under-trained
| 4,032
| 0.00278
| 0.002042
|
155
|
model.layers.22.mlp.gate_proj
| 0.097833
| 4,096
| 14,336
| 3.5
| 6.001974
| -13.491658
| 1.564459
| true
| 0.005651
|
dense
| -13.385668
| -0.836949
| -2.24787
| 64
| 0.145563
| 4,096
| 9
| 4,032
| 1
| 1.667325
| 0.005651
| 25.758545
|
success
| 0.075174
| 0.000001
|
under-trained
| 4,032
| 0.005651
| 0.002382
|
156
|
model.layers.22.mlp.up_proj
| 0.08121
| 4,096
| 14,336
| 3.5
| 6.807247
| -15.786899
| 1.565366
| true
| 0.004796
|
dense
| -15.680063
| -0.863387
| -2.319131
| 64
| 0.136966
| 4,096
| 11
| 4,032
| 1
| 1.750951
| 0.004796
| 28.55909
|
success
| 0.069252
| 0.000001
|
under-trained
| 4,032
| 0.004796
| 0.00219
|
157
|
model.layers.22.self_attn.k_proj
| 0.101224
| 1,024
| 4,096
| 4
| 6.56137
| -19.691763
| 1.135512
| true
| 0.000997
|
dense
| -18.594873
| -1.324471
| -3.001166
| 64
| 0.047373
| 1,024
| 64
| 960
| 1
| 0.695171
| 0.000997
| 47.500237
|
success
| 0.03158
| 0.000001
|
under-trained
| 960
| 0.000997
| 0.000613
|
158
|
model.layers.22.self_attn.o_proj
| 0.064047
| 4,096
| 4,096
| 1
| 14.630623
| -41.505533
| 1.567305
| true
| 0.001456
|
dense
| -41.424312
| -1.22554
| -2.836895
| 64
| 0.059492
| 4,096
| 63
| 4,032
| 1
| 1.717297
| 0.001456
| 40.865242
|
success
| 0.038155
| 0
|
under-trained
| 4,032
| 0.001456
| 0.000861
|
159
|
model.layers.22.self_attn.q_proj
| 0.03411
| 4,096
| 4,096
| 1
| 9.809994
| -26.862289
| 1.565564
| true
| 0.001827
|
dense
| -26.838374
| -1.255902
| -2.738257
| 64
| 0.055475
| 4,096
| 42
| 4,032
| 1
| 1.359412
| 0.001827
| 30.363724
|
success
| 0.042744
| 0
|
under-trained
| 4,032
| 0.001827
| 0.000807
|
160
|
model.layers.22.self_attn.v_proj
| 0.054654
| 1,024
| 4,096
| 4
| 21.498415
| -66.059842
| 1.136995
| true
| 0.000846
|
dense
| -65.872133
| -1.378539
| -3.072777
| 64
| 0.041827
| 1,024
| 20
| 960
| 1
| 4.583585
| 0.000846
| 49.458221
|
success
| 0.029081
| 0.000001
|
under-trained
| 960
| 0.000846
| 0.000671
|
161
|
model.layers.23.mlp.down_proj
| 0.062324
| 4,096
| 14,336
| 3.5
| 27.278771
| -70.935617
| 1.568166
| true
| 0.00251
|
dense
| -70.675347
| -0.879847
| -2.600396
| 64
| 0.131872
| 4,096
| 63
| 4,032
| 1
| 3.310814
| 0.00251
| 52.547184
|
success
| 0.050096
| 0.000001
|
under-trained
| 4,032
| 0.00251
| 0.001983
|
162
|
model.layers.23.mlp.gate_proj
| 0.102704
| 4,096
| 14,336
| 3.5
| 6.167049
| -13.731182
| 1.564025
| true
| 0.005936
|
dense
| -13.667456
| -0.841866
| -2.22654
| 64
| 0.143924
| 4,096
| 9
| 4,032
| 1
| 1.72235
| 0.005936
| 24.247877
|
success
| 0.077042
| 0.000001
|
under-trained
| 4,032
| 0.005936
| 0.002354
|
163
|
model.layers.23.mlp.up_proj
| 0.072697
| 4,096
| 14,336
| 3.5
| 7.470926
| -17.248557
| 1.565169
| true
| 0.004912
|
dense
| -17.20196
| -0.8729
| -2.308758
| 64
| 0.133998
| 4,096
| 13
| 4,032
| 1
| 1.794712
| 0.004912
| 27.280809
|
success
| 0.070084
| 0.000001
|
under-trained
| 4,032
| 0.004912
| 0.00211
|
164
|
model.layers.23.self_attn.k_proj
| 0.050034
| 1,024
| 4,096
| 4
| 9.433142
| -28.276416
| 1.135838
| true
| 0.001006
|
dense
| -27.622812
| -1.349703
| -2.997561
| 64
| 0.044699
| 1,024
| 48
| 960
| 1
| 1.217219
| 0.001006
| 44.448589
|
success
| 0.031712
| 0.000001
|
under-trained
| 960
| 0.001006
| 0.000639
|
165
|
model.layers.23.self_attn.o_proj
| 0.073546
| 4,096
| 4,096
| 1
| 5.384058
| -14.183412
| 1.56385
| true
| 0.002321
|
dense
| -13.935122
| -1.200379
| -2.634335
| 64
| 0.063041
| 4,096
| 12
| 4,032
| 1
| 1.265568
| 0.002321
| 27.161652
|
success
| 0.048176
| 0
| 4,032
| 0.002321
| 0.00099
|
|
166
|
model.layers.23.self_attn.q_proj
| 0.026508
| 4,096
| 4,096
| 1
| 10.00856
| -27.863208
| 1.565898
| true
| 0.001645
|
dense
| -27.811764
| -1.269355
| -2.783938
| 64
| 0.053783
| 4,096
| 49
| 4,032
| 1
| 1.286937
| 0.001645
| 32.702648
|
success
| 0.040554
| 0
|
under-trained
| 4,032
| 0.001645
| 0.000769
|
167
|
model.layers.23.self_attn.v_proj
| 0.066253
| 1,024
| 4,096
| 4
| 17.725424
| -54.425669
| 1.136923
| true
| 0.00085
|
dense
| -53.875092
| -1.360398
| -3.070486
| 64
| 0.043612
| 1,024
| 26
| 960
| 1
| 3.280126
| 0.00085
| 51.296551
|
success
| 0.029158
| 0.000001
|
under-trained
| 960
| 0.00085
| 0.000687
|
168
|
model.layers.24.mlp.down_proj
| 0.088589
| 4,096
| 14,336
| 3.5
| 25.547802
| -65.549568
| 1.568011
| true
| 0.002718
|
dense
| -65.519276
| -0.893975
| -2.565762
| 64
| 0.127651
| 4,096
| 63
| 4,032
| 1
| 3.092732
| 0.002718
| 46.966293
|
success
| 0.052134
| 0.000001
|
under-trained
| 4,032
| 0.002718
| 0.001913
|
169
|
model.layers.24.mlp.gate_proj
| 0.116326
| 4,096
| 14,336
| 3.5
| 5.891471
| -13.039916
| 1.563655
| true
| 0.006119
|
dense
| -12.980163
| -0.84982
| -2.213355
| 64
| 0.141312
| 4,096
| 8
| 4,032
| 1
| 1.729396
| 0.006119
| 23.09589
|
success
| 0.078221
| 0.000001
| 4,032
| 0.006119
| 0.002318
|
|
170
|
model.layers.24.mlp.up_proj
| 0.084695
| 4,096
| 14,336
| 3.5
| 7.723819
| -17.807934
| 1.564945
| true
| 0.004948
|
dense
| -17.779779
| -0.88595
| -2.305587
| 64
| 0.130032
| 4,096
| 14
| 4,032
| 1
| 1.797016
| 0.004948
| 26.280695
|
success
| 0.070341
| 0.000001
|
under-trained
| 4,032
| 0.004948
| 0.002026
|
171
|
model.layers.24.self_attn.k_proj
| 0.042094
| 1,024
| 4,096
| 4
| 9.26434
| -27.923382
| 1.135516
| true
| 0.000968
|
dense
| -27.431276
| -1.393554
| -3.014071
| 64
| 0.040406
| 1,024
| 40
| 960
| 1
| 1.306707
| 0.000968
| 41.736637
|
success
| 0.031115
| 0.000001
|
under-trained
| 960
| 0.000968
| 0.000592
|
172
|
model.layers.24.self_attn.o_proj
| 0.063429
| 4,096
| 4,096
| 1
| 18.543035
| -53.72646
| 1.567791
| true
| 0.001267
|
dense
| -53.706925
| -1.265283
| -2.897393
| 64
| 0.05429
| 4,096
| 58
| 4,032
| 1
| 2.303513
| 0.001267
| 42.865685
|
success
| 0.035588
| 0
|
under-trained
| 4,032
| 0.001267
| 0.000804
|
173
|
model.layers.24.self_attn.q_proj
| 0.03856
| 4,096
| 4,096
| 1
| 10.280278
| -28.986392
| 1.565957
| true
| 0.001515
|
dense
| -28.963955
| -1.323311
| -2.819612
| 64
| 0.0475
| 4,096
| 49
| 4,032
| 1
| 1.325754
| 0.001515
| 31.354559
|
success
| 0.038922
| 0
|
under-trained
| 4,032
| 0.001515
| 0.000681
|
174
|
model.layers.24.self_attn.v_proj
| 0.059587
| 1,024
| 4,096
| 4
| 26.995273
| -83.393638
| 1.137081
| true
| 0.000814
|
dense
| -83.276914
| -1.383491
| -3.089194
| 64
| 0.041353
| 1,024
| 16
| 960
| 1
| 6.498818
| 0.000814
| 50.781246
|
success
| 0.028537
| 0.000001
|
under-trained
| 960
| 0.000814
| 0.000675
|
175
|
model.layers.25.mlp.down_proj
| 0.088452
| 4,096
| 14,336
| 3.5
| 22.674786
| -57.540608
| 1.567786
| true
| 0.0029
|
dense
| -57.530036
| -0.90192
| -2.537647
| 64
| 0.125337
| 4,096
| 64
| 4,032
| 1
| 2.709348
| 0.0029
| 43.224232
|
success
| 0.053849
| 0.000001
|
under-trained
| 4,032
| 0.0029
| 0.001866
|
176
|
model.layers.25.mlp.gate_proj
| 0.115405
| 4,096
| 14,336
| 3.5
| 7.173909
| -15.938461
| 1.563881
| true
| 0.006002
|
dense
| -15.919066
| -0.851489
| -2.221726
| 64
| 0.14077
| 4,096
| 12
| 4,032
| 1
| 1.782254
| 0.006002
| 23.455097
|
success
| 0.077471
| 0.000001
|
under-trained
| 4,032
| 0.006002
| 0.002222
|
177
|
model.layers.25.mlp.up_proj
| 0.117501
| 4,096
| 14,336
| 3.5
| 7.724192
| -17.831249
| 1.564917
| true
| 0.004915
|
dense
| -17.805651
| -0.893028
| -2.308494
| 64
| 0.12793
| 4,096
| 13
| 4,032
| 1
| 1.864955
| 0.004915
| 26.029451
|
success
| 0.070106
| 0.000001
|
under-trained
| 4,032
| 0.004915
| 0.002006
|
178
|
model.layers.25.self_attn.k_proj
| 0.102301
| 1,024
| 4,096
| 4
| 7.238527
| -22.14815
| 1.135898
| true
| 0.000871
|
dense
| -21.109452
| -1.379533
| -3.059759
| 64
| 0.041732
| 1,024
| 62
| 960
| 1
| 0.792294
| 0.000871
| 47.887905
|
success
| 0.02952
| 0.000001
|
under-trained
| 960
| 0.000871
| 0.000555
|
179
|
model.layers.25.self_attn.o_proj
| 0.072905
| 4,096
| 4,096
| 1
| 15.151347
| -43.711322
| 1.567377
| true
| 0.001303
|
dense
| -43.662192
| -1.280084
| -2.884979
| 64
| 0.052471
| 4,096
| 63
| 4,032
| 1
| 1.782902
| 0.001303
| 40.262028
|
success
| 0.0361
| 0
|
under-trained
| 4,032
| 0.001303
| 0.000762
|
180
|
model.layers.25.self_attn.q_proj
| 0.024015
| 4,096
| 4,096
| 1
| 10.113745
| -28.551966
| 1.565922
| true
| 0.001503
|
dense
| -28.520994
| -1.319469
| -2.823085
| 64
| 0.047922
| 4,096
| 40
| 4,032
| 1
| 1.44101
| 0.001503
| 31.887217
|
success
| 0.038767
| 0
|
under-trained
| 4,032
| 0.001503
| 0.000703
|
181
|
model.layers.25.self_attn.v_proj
| 0.050871
| 1,024
| 4,096
| 4
| 22.102889
| -67.771573
| 1.13697
| true
| 0.000859
|
dense
| -67.671703
| -1.382476
| -3.066186
| 64
| 0.04145
| 1,024
| 20
| 960
| 1
| 4.718749
| 0.000859
| 48.273659
|
success
| 0.029303
| 0.000001
|
under-trained
| 960
| 0.000859
| 0.000666
|
182
|
model.layers.26.mlp.down_proj
| 0.099921
| 4,096
| 14,336
| 3.5
| 21.817002
| -54.155571
| 1.567389
| true
| 0.003294
|
dense
| -54.155059
| -0.910668
| -2.482265
| 64
| 0.122838
| 4,096
| 64
| 4,032
| 1
| 2.602125
| 0.003294
| 37.290382
|
success
| 0.057394
| 0.000001
|
under-trained
| 4,032
| 0.003294
| 0.001823
|
183
|
model.layers.26.mlp.gate_proj
| 0.08999
| 4,096
| 14,336
| 3.5
| 6.457537
| -14.237801
| 1.563379
| true
| 0.00624
|
dense
| -14.209257
| -0.85408
| -2.204835
| 64
| 0.139933
| 4,096
| 11
| 4,032
| 1
| 1.645509
| 0.00624
| 22.426134
|
success
| 0.078992
| 0.000001
|
under-trained
| 4,032
| 0.00624
| 0.002202
|
184
|
model.layers.26.mlp.up_proj
| 0.100135
| 4,096
| 14,336
| 3.5
| 6.472845
| -14.783802
| 1.564256
| true
| 0.0052
|
dense
| -14.732363
| -0.893702
| -2.283973
| 64
| 0.127732
| 4,096
| 11
| 4,032
| 1
| 1.650125
| 0.0052
| 24.562418
|
success
| 0.072113
| 0.000001
|
under-trained
| 4,032
| 0.0052
| 0.002022
|
185
|
model.layers.26.self_attn.k_proj
| 0.030124
| 1,024
| 4,096
| 4
| 8.838954
| -26.360515
| 1.135024
| true
| 0.001042
|
dense
| -26.077297
| -1.411301
| -2.982311
| 64
| 0.038788
| 1,024
| 34
| 960
| 1
| 1.34437
| 0.001042
| 37.240086
|
success
| 0.032273
| 0.000001
|
under-trained
| 960
| 0.001042
| 0.000578
|
186
|
model.layers.26.self_attn.o_proj
| 0.080145
| 4,096
| 4,096
| 1
| 4.237009
| -11.025414
| 1.560244
| true
| 0.002499
|
dense
| -10.726383
| -1.24904
| -2.602169
| 64
| 0.056359
| 4,096
| 12
| 4,032
| 1
| 0.934444
| 0.002499
| 22.549124
|
success
| 0.049994
| 0
| 4,032
| 0.002499
| 0.000862
|
|
187
|
model.layers.26.self_attn.q_proj
| 0.036869
| 4,096
| 4,096
| 1
| 8.40632
| -23.18886
| 1.564288
| true
| 0.001744
|
dense
| -23.157452
| -1.324999
| -2.758503
| 64
| 0.047315
| 4,096
| 64
| 4,032
| 1
| 0.92579
| 0.001744
| 27.133411
|
success
| 0.041759
| 0
|
under-trained
| 4,032
| 0.001744
| 0.000637
|
188
|
model.layers.26.self_attn.v_proj
| 0.06321
| 1,024
| 4,096
| 4
| 14.601195
| -43.401385
| 1.136401
| true
| 0.001065
|
dense
| -43.379188
| -1.386946
| -2.972454
| 64
| 0.041025
| 1,024
| 38
| 960
| 1
| 2.206405
| 0.001065
| 38.5042
|
success
| 0.032642
| 0.000001
|
under-trained
| 960
| 0.001065
| 0.00062
|
189
|
model.layers.27.mlp.down_proj
| 0.125634
| 4,096
| 14,336
| 3.5
| 20.134975
| -48.749573
| 1.566567
| true
| 0.003792
|
dense
| -48.749521
| -0.922233
| -2.421139
| 64
| 0.11961
| 4,096
| 64
| 4,032
| 1
| 2.391872
| 0.003792
| 31.543194
|
success
| 0.061579
| 0.000001
|
under-trained
| 4,032
| 0.003792
| 0.001763
|
190
|
model.layers.27.mlp.gate_proj
| 0.080541
| 4,096
| 14,336
| 3.5
| 5.975536
| -12.835613
| 1.561437
| true
| 0.007112
|
dense
| -12.813628
| -0.853734
| -2.148027
| 64
| 0.140044
| 4,096
| 12
| 4,032
| 1
| 1.436314
| 0.007112
| 19.692146
|
success
| 0.084331
| 0.000001
| 4,032
| 0.007112
| 0.002181
|
|
191
|
model.layers.27.mlp.up_proj
| 0.090913
| 4,096
| 14,336
| 3.5
| 5.596395
| -12.553345
| 1.56291
| true
| 0.005713
|
dense
| -12.483229
| -0.893531
| -2.243113
| 64
| 0.127782
| 4,096
| 10
| 4,032
| 1
| 1.453508
| 0.005713
| 22.365685
|
success
| 0.075586
| 0.000001
| 4,032
| 0.005713
| 0.002055
|
|
192
|
model.layers.27.self_attn.k_proj
| 0.061444
| 1,024
| 4,096
| 4
| 10.763951
| -32.338098
| 1.135765
| true
| 0.00099
|
dense
| -32.04649
| -1.391835
| -3.004296
| 64
| 0.040566
| 1,024
| 27
| 960
| 1
| 1.879073
| 0.00099
| 40.969563
|
success
| 0.031467
| 0.000001
|
under-trained
| 960
| 0.00099
| 0.000631
|
193
|
model.layers.27.self_attn.o_proj
| 0.09495
| 4,096
| 4,096
| 1
| 11.720985
| -32.742357
| 1.565708
| true
| 0.001609
|
dense
| -32.718908
| -1.288713
| -2.793482
| 64
| 0.051438
| 4,096
| 64
| 4,032
| 1
| 1.340123
| 0.001609
| 31.971897
|
success
| 0.040111
| 0
|
under-trained
| 4,032
| 0.001609
| 0.000725
|
194
|
model.layers.27.self_attn.q_proj
| 0.030258
| 4,096
| 4,096
| 1
| 9.312165
| -26.042982
| 1.565224
| true
| 0.001597
|
dense
| -26.018858
| -1.331789
| -2.796662
| 64
| 0.046581
| 4,096
| 51
| 4,032
| 1
| 1.163936
| 0.001597
| 29.165789
|
success
| 0.039964
| 0
|
under-trained
| 4,032
| 0.001597
| 0.000656
|
195
|
model.layers.27.self_attn.v_proj
| 0.054073
| 1,024
| 4,096
| 4
| 16.699773
| -50.638051
| 1.136774
| true
| 0.000928
|
dense
| -50.55997
| -1.387949
| -3.03226
| 64
| 0.040931
| 1,024
| 31
| 960
| 1
| 2.819762
| 0.000928
| 44.087074
|
success
| 0.03047
| 0.000001
|
under-trained
| 960
| 0.000928
| 0.000633
|
196
|
model.layers.28.mlp.down_proj
| 0.147856
| 4,096
| 14,336
| 3.5
| 17.659206
| -40.82923
| 1.564656
| true
| 0.004875
|
dense
| -40.829228
| -0.920379
| -2.312065
| 64
| 0.120122
| 4,096
| 64
| 4,032
| 1
| 2.082401
| 0.004875
| 24.642588
|
success
| 0.069818
| 0.000001
|
under-trained
| 4,032
| 0.004875
| 0.001747
|
197
|
model.layers.28.mlp.gate_proj
| 0.091559
| 4,096
| 14,336
| 3.5
| 5.433214
| -11.60878
| 1.560451
| true
| 0.007301
|
dense
| -11.576672
| -0.863321
| -2.136632
| 64
| 0.136987
| 4,096
| 11
| 4,032
| 1
| 1.336664
| 0.007301
| 18.763388
|
success
| 0.085444
| 0.000001
| 4,032
| 0.007301
| 0.002153
|
|
198
|
model.layers.28.mlp.up_proj
| 0.081664
| 4,096
| 14,336
| 3.5
| 5.483033
| -12.276219
| 1.562416
| true
| 0.005768
|
dense
| -12.196294
| -0.893248
| -2.238947
| 64
| 0.127865
| 4,096
| 11
| 4,032
| 1
| 1.351685
| 0.005768
| 22.166574
|
success
| 0.07595
| 0.000001
| 4,032
| 0.005768
| 0.002042
|
|
199
|
model.layers.28.self_attn.k_proj
| 0.049955
| 1,024
| 4,096
| 4
| 8.914576
| -27.248532
| 1.135713
| true
| 0.000878
|
dense
| -26.40493
| -1.391175
| -3.056627
| 64
| 0.040628
| 1,024
| 45
| 960
| 1
| 1.179835
| 0.000878
| 46.28624
|
success
| 0.029627
| 0.000001
|
under-trained
| 960
| 0.000878
| 0.000583
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.