TOP ACTIVATIONS
MAX = 0.708

I â m sorry Jane . I
those in need â even <|endoftext|> One day
It â s almost just as beautiful
It â s a magic book !
That â s a special necklace ,
't know the rh in oc eros was wild . Can
Let â s go find something else
! â â
That â s the gas from the
That â s right , Emma .
by one 1 , 2 ,
â The small child thanked the
Let â s play a joke .
Let â s not think about the
Let â s have a stretching race
Let â s explore and have some
Who â s there ? â
Please don â t eat me , â
ash tr ay ! He brought it inside ,
here ? â The box answered ,

MIN ACTIVATIONS
MIN = -0.367

always make things work out perfectly . Sometimes , things just
home and then they sne aked in . The
, the band was playing perfectly ! The animals cheered and
wanted something that fit just perfectly . Then Tommy went to
" This swing fits me perfectly , Mom !" Lily said
the skirt and it fit perfectly . She was so happy
feeling sleepy . He yawn ed , rubbing his eyes .
. They sne aked into their room and took
someone who would match him perfectly . After a while ,
The clown smiled and wink ed . He took out another
his car . It fit perfectly . He was happy .
to them . It sniff ed them . It l icked
. And he never yawn ed while singing his green song
stay there . She fit perfectly into the chair .
heet . He did it perfectly ! The tutor
him and it fit him perfectly . He looked in the
compassionate and kept the rhythm perfectly . She was so proud
found one that fit him perfectly and it was dry ,
and forgot all about her jealousy . After dinner
on a bench and peek ed inside . "

INTERVAL 0.439 - 0.708
CONTAINS 1.209%

my , Daddy , help ! I lost my marble !"
tree out of the ground . Mike was so happy and
home to play . So she asked her mom
want to leave . The rain started to fall
that were all mixed up . Jane didn 't know what

INTERVAL 0.171 - 0.439
CONTAINS 17.440%

shouted with happiness and the p anda felt so proud .
Do you need help ?" Cotton said , " Yes ,
, but G ogo dodged . He used his horns to
smiled , " Yes I do . Come , and I
didn 't work anymore and she felt sad .

INTERVAL -0.098 - 0.171
CONTAINS 46.306%

". The man agreed . Together , the duck
owl thought this was a good idea and the two friends
it was only fair . So , they decided to let
inside . As they stepped over the threshold they heard a
buddy ! Do you want to come take a closer look

INTERVAL -0.367 - -0.098
CONTAINS 35.044%

and looked up . There was a fox ! The fox
out and jumped onto Matt 's shoulders . The monkey stayed
a little boy saw it . The little boy was only
played in the park and laughed a lot . They played
put on their life jackets and w aded out into the
Token
Token ID129
Feature activation+6.69e-01

Pos logit contributions
 faded+0.332
 lighter+0.319
mail+0.319
Lost+0.316
 tangled+0.313
Neg logit contributions
 how-0.253
ap-0.232
iu-0.223
 why-0.216
ext-0.209
Token
Token ID241
Feature activation+5.96e-01

Pos logit contributions
 interests+0.680
 Amen+0.672
iu+0.644
ric+0.627
pieces+0.576
Neg logit contributions
 faded-0.711
 tangled-0.710
 healed-0.704
 melted-0.689
chy-0.686
TokenI
Token ID40
Feature activation+4.90e-01

Pos logit contributions
iu+0.688
 Amen+0.657
ric+0.620
ext+0.620
icol+0.618
Neg logit contributions
 faded-0.616
 tangled-0.610
 shone-0.583
 melted-0.581
 flowed-0.559
Tokenâ
Token ID22940
Feature activation+5.02e-01

Pos logit contributions
 Amen+0.382
ric+0.351
 interests+0.334
iu+0.319
ext+0.293
Neg logit contributions
 faded-0.691
mail-0.659
 tangled-0.654
 flowed-0.653
Factory-0.650
Token
Token ID26391
Feature activation+5.15e-01

Pos logit contributions
iu+0.427
 Amen+0.416
ric+0.415
ext+0.391
 interests+0.363
Neg logit contributions
 faded-0.682
 tangled-0.648
 melted-0.647
 shone-0.642
 healed-0.630
Token
Token ID8151
Feature activation+5.30e-01

Pos logit contributions
iu+0.500
 Amen+0.491
ric+0.467
ext+0.452
 interests+0.427
Neg logit contributions
 faded-0.621
 tangled-0.599
 healed-0.592
 shone-0.591
 melted-0.590
Tokenm
Token ID76
Feature activation+4.21e-01

Pos logit contributions
ric+0.571
 Amen+0.476
ext+0.470
iu+0.455
 interests+0.453
Neg logit contributions
 faded-0.646
 shone-0.592
 melted-0.590
 flowed-0.587
 iv-0.577
Token sorry
Token ID7926
Feature activation+4.08e-01

Pos logit contributions
ric+0.498
 Amen+0.442
iu+0.439
ext+0.420
ap+0.406
Neg logit contributions
 faded-0.497
mail-0.467
 shone-0.445
 melted-0.434
 tangled-0.432
Token Jane
Token ID12091
Feature activation+3.40e-01

Pos logit contributions
iu+0.395
ric+0.382
 Amen+0.382
ext+0.359
 interests+0.329
Neg logit contributions
 faded-0.516
 melted-0.484
 tangled-0.484
 shone-0.481
chy-0.468
Token.
Token ID13
Feature activation+4.64e-01

Pos logit contributions
 Amen+0.247
ric+0.232
!'+0.231
iu+0.219
 how+0.218
Neg logit contributions
Factory-0.468
mail-0.466
 darker-0.465
Ancient-0.464
 flowed-0.462
Token I
Token ID314
Feature activation+3.65e-01

Pos logit contributions
 Amen+0.330
ric+0.315
iu+0.313
ext+0.271
 interests+0.257
Neg logit contributions
 faded-0.715
 melted-0.685
 tangled-0.675
 shone-0.672
chy-0.658
Token those
Token ID883
Feature activation+4.95e-01

Pos logit contributions
 faded+0.148
 healed+0.136
 shone+0.136
 melted+0.136
mail+0.136
Neg logit contributions
iu-0.157
 Amen-0.149
ric-0.146
ext-0.138
 interests-0.135
Token in
Token ID287
Feature activation+4.53e-01

Pos logit contributions
iu+0.496
ric+0.494
 Amen+0.480
ext+0.449
 interests+0.445
Neg logit contributions
 faded-0.597
 tangled-0.559
 shone-0.550
 melted-0.550
 healed-0.538
Token need
Token ID761
Feature activation+4.37e-01

Pos logit contributions
ric+0.394
 Amen+0.373
 Wed+0.361
 how+0.355
 interests+0.352
Neg logit contributions
 faded-0.597
 shone-0.561
chy-0.557
 flowed-0.556
mail-0.555
Tokenâ
Token ID22940
Feature activation+4.69e-01

Pos logit contributions
 how+0.290
 Amen+0.274
 interests+0.271
ric+0.271
 Wed+0.263
Neg logit contributions
 faded-0.645
 shone-0.612
 flowed-0.611
 tangled-0.610
chy-0.609
Token
Token ID26391
Feature activation+4.83e-01

Pos logit contributions
iu+0.411
 Amen+0.404
ric+0.397
ext+0.353
 interests+0.340
Neg logit contributions
 faded-0.646
 shone-0.610
 tangled-0.608
 melted-0.598
 disappeared-0.585
Token
Token ID447
Feature activation+5.27e-01

Pos logit contributions
iu+0.432
 Amen+0.427
ric+0.411
ext+0.382
 interests+0.362
Neg logit contributions
 faded-0.633
 tangled-0.601
 melted-0.597
 shone-0.590
 healed-0.585
Token
Token ID251
Feature activation+4.77e-01

Pos logit contributions
 Amen+0.169
ric+0.116
!'+0.080
'.+0.053
 We+0.051
Neg logit contributions
 faded-1.046
 prick-1.016
 iv-0.999
 flowed-0.996
 leaked-0.986
Tokeneven
Token ID10197
Feature activation+3.11e-01

Pos logit contributions
 Amen+0.376
iu+0.353
ric+0.330
ext+0.314
 Wed+0.308
Neg logit contributions
 faded-0.692
 melted-0.665
 tangled-0.657
 shone-0.655
chy-0.650
Token<|endoftext|>
Token ID50256
Feature activation+4.17e-01

Pos logit contributions
iu+0.345
 Amen+0.341
 interests+0.329
 how+0.323
ric+0.322
Neg logit contributions
 faded-0.315
 flowed-0.301
 melted-0.299
 shone-0.299
mail-0.298
TokenOne
Token ID3198
Feature activation+4.09e-01

Pos logit contributions
ric+0.489
 Amen+0.483
iu+0.472
ext+0.453
 Wed+0.432
Neg logit contributions
 faded-0.433
 melted-0.397
 tangled-0.388
chy-0.382
 stiff-0.379
Token day
Token ID1110
Feature activation+3.29e-01

Pos logit contributions
ric+0.371
 Amen+0.335
 how+0.335
 Wed+0.335
 interests+0.330
Neg logit contributions
 faded-0.512
mail-0.505
 iv-0.493
 tangled-0.487
Factory-0.486
Token
Token ID129
Feature activation+6.69e-01

Pos logit contributions
 faded+0.146
 tangled+0.137
 melted+0.137
 shone+0.136
 healed+0.132
Neg logit contributions
iu-0.114
 Amen-0.111
ric-0.110
ext-0.102
 interests-0.098
Token
Token ID241
Feature activation+5.96e-01

Pos logit contributions
 interests+0.680
 Amen+0.672
iu+0.644
ric+0.627
pieces+0.576
Neg logit contributions
 faded-0.711
 tangled-0.710
 healed-0.704
 melted-0.689
chy-0.686
TokenIt
Token ID1026
Feature activation+4.49e-01

Pos logit contributions
iu+0.688
 Amen+0.657
ric+0.620
ext+0.620
icol+0.618
Neg logit contributions
 faded-0.616
 tangled-0.610
 shone-0.583
 melted-0.581
 flowed-0.559
Tokenâ
Token ID22940
Feature activation+4.98e-01

Pos logit contributions
ric+0.284
 how+0.258
 interests+0.233
 Amen+0.233
 We+0.225
Neg logit contributions
 faded-0.708
mail-0.671
 tangled-0.651
Factory-0.636
 dim-0.635
Token
Token ID26391
Feature activation+5.12e-01

Pos logit contributions
iu+0.425
 Amen+0.418
ric+0.412
ext+0.380
 interests+0.356
Neg logit contributions
 faded-0.682
 tangled-0.645
 melted-0.642
 shone-0.641
 healed-0.620
Token
Token ID8151
Feature activation+5.27e-01

Pos logit contributions
iu+0.460
 Amen+0.456
ric+0.439
ext+0.417
 interests+0.389
Neg logit contributions
 faded-0.667
 tangled-0.633
 melted-0.630
 shone-0.629
 healed-0.616
Tokens
Token ID82
Feature activation+4.54e-01

Pos logit contributions
ric+0.564
ext+0.436
!'+0.422
 Amen+0.420
 interests+0.414
Neg logit contributions
 faded-0.668
 fairy-0.619
 iv-0.612
 shone-0.606
 prick-0.604
Token almost
Token ID2048
Feature activation+3.97e-01

Pos logit contributions
ric+0.412
 Amen+0.359
iu+0.343
 how+0.337
 interests+0.331
Neg logit contributions
 faded-0.617
mail-0.586
 shone-0.573
 tangled-0.572
 flowed-0.557
Token just
Token ID655
Feature activation+4.09e-01

Pos logit contributions
ric+0.323
iu+0.299
 how+0.298
 Amen+0.286
 you+0.280
Neg logit contributions
mail-0.525
Factory-0.512
Ancient-0.505
 faded-0.495
Bee-0.493
Token as
Token ID355
Feature activation+3.96e-01

Pos logit contributions
ric+0.337
 how+0.329
iu+0.307
 Amen+0.301
 interests+0.296
Neg logit contributions
mail-0.545
Factory-0.534
Ancient-0.526
 faded-0.526
 tangled-0.512
Token beautiful
Token ID4950
Feature activation+3.35e-01

Pos logit contributions
 you+0.392
 how+0.354
 we+0.340
 Wed+0.332
 our+0.323
Neg logit contributions
 faded-0.462
mail-0.461
Factory-0.458
 iv-0.451
 triggers-0.440
Token
Token ID129
Feature activation+6.69e-01

Pos logit contributions
 faded+0.221
 tangled+0.207
chy+0.207
 flowed+0.206
 melted+0.203
Neg logit contributions
iu-0.351
 Amen-0.349
ric-0.323
 interests-0.316
ext-0.311
Token
Token ID241
Feature activation+5.96e-01

Pos logit contributions
 interests+0.680
 Amen+0.672
iu+0.644
ric+0.627
pieces+0.576
Neg logit contributions
 faded-0.711
 tangled-0.710
 healed-0.704
 melted-0.689
chy-0.686
TokenIt
Token ID1026
Feature activation+4.49e-01

Pos logit contributions
iu+0.688
 Amen+0.657
ric+0.620
ext+0.620
icol+0.618
Neg logit contributions
 faded-0.616
 tangled-0.610
 shone-0.583
 melted-0.581
 flowed-0.559
Tokenâ
Token ID22940
Feature activation+4.98e-01

Pos logit contributions
ric+0.284
 how+0.258
 interests+0.233
 Amen+0.233
 We+0.225
Neg logit contributions
 faded-0.708
mail-0.671
 tangled-0.651
Factory-0.636
 dim-0.635
Token
Token ID26391
Feature activation+5.12e-01

Pos logit contributions
iu+0.425
 Amen+0.418
ric+0.412
ext+0.380
 interests+0.356
Neg logit contributions
 faded-0.682
 tangled-0.645
 melted-0.642
 shone-0.641
 healed-0.620
Token
Token ID8151
Feature activation+5.27e-01

Pos logit contributions
iu+0.460
 Amen+0.456
ric+0.439
ext+0.417
 interests+0.389
Neg logit contributions
 faded-0.667
 tangled-0.633
 melted-0.630
 shone-0.629
 healed-0.616
Tokens
Token ID82
Feature activation+4.54e-01

Pos logit contributions
ric+0.564
ext+0.436
!'+0.422
 Amen+0.420
 interests+0.414
Neg logit contributions
 faded-0.668
 fairy-0.619
 iv-0.612
 shone-0.606
 prick-0.604
Token a
Token ID257
Feature activation+3.82e-01

Pos logit contributions
ric+0.412
 Amen+0.359
iu+0.343
 how+0.337
 interests+0.331
Neg logit contributions
 faded-0.617
mail-0.586
 shone-0.573
 tangled-0.572
 flowed-0.557
Token magic
Token ID5536
Feature activation+3.34e-01

Pos logit contributions
 Amen+0.449
ric+0.415
iu+0.400
 Wed+0.381
 interests+0.380
Neg logit contributions
 faded-0.378
mail-0.376
 flowed-0.364
 tangled-0.360
Factory-0.354
Token book
Token ID1492
Feature activation+3.29e-01

Pos logit contributions
ric+0.350
 Amen+0.348
iu+0.342
 interests+0.340
 how+0.332
Neg logit contributions
 tangled-0.349
 faded-0.330
mail-0.319
 crushed-0.314
 melted-0.313
Token!
Token ID0
Feature activation+4.40e-01

Pos logit contributions
 how+0.164
 where+0.143
 you+0.132
 about+0.126
 We+0.122
Neg logit contributions
 faded-0.534
 tangled-0.526
Bee-0.526
 melted-0.525
 weaker-0.524
Token
Token ID129
Feature activation+6.69e-01

Pos logit contributions
 faded+0.338
 tangled+0.319
 melted+0.317
 shone+0.316
 healed+0.307
Neg logit contributions
iu-0.252
 Amen-0.247
ric-0.245
ext-0.230
 interests-0.216
Token
Token ID241
Feature activation+5.96e-01

Pos logit contributions
 interests+0.680
 Amen+0.672
iu+0.644
ric+0.627
pieces+0.576
Neg logit contributions
 faded-0.711
 tangled-0.710
 healed-0.704
 melted-0.689
chy-0.686
TokenThat
Token ID2504
Feature activation+4.35e-01

Pos logit contributions
iu+0.688
 Amen+0.657
ric+0.620
ext+0.620
icol+0.618
Neg logit contributions
 faded-0.616
 tangled-0.610
 shone-0.583
 melted-0.581
 flowed-0.559
Tokenâ
Token ID22940
Feature activation+4.97e-01

Pos logit contributions
ric+0.390
iu+0.362
 Amen+0.361
 interests+0.341
 Wed+0.336
Neg logit contributions
 faded-0.573
mail-0.529
 tangled-0.526
 melted-0.519
 shone-0.519
Token
Token ID26391
Feature activation+5.10e-01

Pos logit contributions
iu+0.375
 Amen+0.365
ric+0.361
ext+0.326
 interests+0.305
Neg logit contributions
 faded-0.733
 tangled-0.693
 shone-0.691
 melted-0.690
 flowed-0.668
Token
Token ID8151
Feature activation+5.26e-01

Pos logit contributions
iu+0.474
 Amen+0.473
ric+0.450
ext+0.433
 interests+0.404
Neg logit contributions
 faded-0.642
 tangled-0.611
 melted-0.609
 shone-0.606
 healed-0.597
Tokens
Token ID82
Feature activation+4.53e-01

Pos logit contributions
ric+0.568
ext+0.444
 interests+0.438
!'+0.437
 Amen+0.427
Neg logit contributions
 faded-0.659
 fairy-0.611
 iv-0.596
 prick-0.595
 shone-0.592
Token a
Token ID257
Feature activation+3.80e-01

Pos logit contributions
ric+0.399
iu+0.384
 Amen+0.381
ext+0.349
 interests+0.338
Neg logit contributions
 faded-0.616
 tangled-0.575
 shone-0.573
 melted-0.570
mail-0.556
Token special
Token ID2041
Feature activation+3.64e-01

Pos logit contributions
 Amen+0.444
ric+0.411
iu+0.408
 interests+0.386
 Wed+0.386
Neg logit contributions
 faded-0.385
mail-0.369
 tangled-0.362
 flowed-0.358
 iv-0.344
Token necklace
Token ID38987
Feature activation+3.30e-01

Pos logit contributions
 Amen+0.396
iu+0.380
ric+0.380
 interests+0.367
 Wed+0.355
Neg logit contributions
 faded-0.395
 tangled-0.382
 flowed-0.358
 healed-0.356
 melted-0.355
Token,
Token ID11
Feature activation+4.19e-01

Pos logit contributions
 how+0.192
iu+0.187
 Amen+0.186
ric+0.178
 interests+0.170
Neg logit contributions
 faded-0.499
 tangled-0.477
 stiff-0.468
Factory-0.466
 flowed-0.465
Token't
Token ID470
Feature activation+5.35e-01

Pos logit contributions
 faded+0.321
 tangled+0.311
chy+0.306
mail+0.306
 shone+0.303
Neg logit contributions
 how-0.135
iu-0.134
 Amen-0.130
 why-0.116
ric-0.111
Token know
Token ID760
Feature activation+4.24e-01

Pos logit contributions
iu+0.506
ric+0.497
 Amen+0.493
ext+0.461
icol+0.421
Neg logit contributions
 faded-0.693
 melted-0.652
 tangled-0.650
 shone-0.637
 healed-0.633
Token the
Token ID262
Feature activation+4.13e-01

Pos logit contributions
iu+0.202
ric+0.199
 Amen+0.193
 how+0.182
ext+0.172
Neg logit contributions
 faded-0.721
 tangled-0.696
 melted-0.686
 shone-0.685
mail-0.674
Token rh
Token ID9529
Feature activation+4.57e-01

Pos logit contributions
ric+0.499
 Amen+0.497
iu+0.490
 interests+0.457
 Wed+0.429
Neg logit contributions
 faded-0.397
 flowed-0.360
mail-0.353
 shone-0.350
 tangled-0.348
Tokenin
Token ID259
Feature activation+3.60e-01

Pos logit contributions
iu+0.652
 Amen+0.647
ric+0.643
ext+0.614
 interests+0.591
Neg logit contributions
 faded-0.362
 tangled-0.327
 shone-0.323
 melted-0.323
 flowed-0.307
Tokenoc
Token ID420
Feature activation+5.21e-01

Pos logit contributions
 Amen+0.186
iu+0.165
ric+0.161
 how+0.138
 interests+0.120
Neg logit contributions
 faded-0.616
 tangled-0.611
 flowed-0.599
 darker-0.588
 stiff-0.588
Tokeneros
Token ID27498
Feature activation+3.49e-01

Pos logit contributions
 Amen+0.559
iu+0.557
ric+0.544
ext+0.531
 interests+0.503
Neg logit contributions
 faded-0.576
 melted-0.548
 tangled-0.537
 shone-0.537
 healed-0.524
Token was
Token ID373
Feature activation+3.64e-01

Pos logit contributions
 how+0.305
 why+0.280
 what+0.253
 you+0.247
 where+0.244
Neg logit contributions
 stiff-0.466
 tangled-0.456
 faded-0.444
 dim-0.435
 darker-0.433
Token wild
Token ID4295
Feature activation+3.28e-01

Pos logit contributions
 how+0.341
 Amen+0.319
ric+0.317
 where+0.307
 why+0.306
Neg logit contributions
Factory-0.439
Ancient-0.436
mail-0.427
Bee-0.419
 flowed-0.417
Token.
Token ID13
Feature activation+4.48e-01

Pos logit contributions
iu+0.283
 Amen+0.280
ric+0.277
ext+0.255
 interests+0.243
Neg logit contributions
 faded-0.438
 tangled-0.414
 melted-0.410
 shone-0.410
 healed-0.400
Token Can
Token ID1680
Feature activation+3.67e-01

Pos logit contributions
 Amen+0.353
ric+0.324
iu+0.303
 We+0.296
ext+0.288
Neg logit contributions
 faded-0.675
 melted-0.670
 shone-0.651
 tangled-0.648
chy-0.635
Token
Token ID129
Feature activation+6.69e-01

Pos logit contributions
 faded+0.334
 flowed+0.330
 tangled+0.324
 bothered+0.322
 thinner+0.318
Neg logit contributions
 Amen-0.155
 Wed-0.139
ossible-0.119
iu-0.109
ext-0.105
Token
Token ID241
Feature activation+5.96e-01

Pos logit contributions
 interests+0.680
 Amen+0.672
iu+0.644
ric+0.627
pieces+0.576
Neg logit contributions
 faded-0.711
 tangled-0.710
 healed-0.704
 melted-0.689
chy-0.686
TokenLet
Token ID5756
Feature activation+4.15e-01

Pos logit contributions
iu+0.688
 Amen+0.657
ric+0.620
ext+0.620
icol+0.618
Neg logit contributions
 faded-0.616
 tangled-0.610
 shone-0.583
 melted-0.581
 flowed-0.559
Tokenâ
Token ID22940
Feature activation+4.90e-01

Pos logit contributions
 Amen+0.375
iu+0.372
ric+0.370
ext+0.333
 interests+0.300
Neg logit contributions
 faded-0.554
 melted-0.537
 tangled-0.528
 shone-0.523
 healed-0.518
Token
Token ID26391
Feature activation+5.05e-01

Pos logit contributions
iu+0.420
 Amen+0.414
ric+0.404
ext+0.402
 interests+0.366
Neg logit contributions
 faded-0.637
 melted-0.613
 tangled-0.610
 healed-0.606
 shone-0.599
Token
Token ID8151
Feature activation+5.21e-01

Pos logit contributions
iu+0.537
 Amen+0.533
ric+0.482
ext+0.477
 interests+0.457
Neg logit contributions
 faded-0.544
 healed-0.538
 tangled-0.528
 melted-0.519
 shone-0.518
Tokens
Token ID82
Feature activation+4.48e-01

Pos logit contributions
ric+0.591
 Amen+0.544
iu+0.543
ext+0.527
 interests+0.501
Neg logit contributions
 faded-0.583
 melted-0.533
 shone-0.531
 tangled-0.526
 flowed-0.513
Token go
Token ID467
Feature activation+3.90e-01

Pos logit contributions
iu+0.417
 Amen+0.405
ric+0.383
ext+0.373
icol+0.334
Neg logit contributions
 faded-0.584
 melted-0.559
 tangled-0.557
 shone-0.555
 healed-0.544
Token find
Token ID1064
Feature activation+3.45e-01

Pos logit contributions
iu+0.387
 Amen+0.374
ric+0.370
ext+0.352
 interests+0.329
Neg logit contributions
 faded-0.481
 tangled-0.462
 melted-0.459
 shone-0.455
 healed-0.443
Token something
Token ID1223
Feature activation+3.36e-01

Pos logit contributions
iu+0.288
 Amen+0.278
ric+0.274
 how+0.266
ext+0.262
Neg logit contributions
 faded-0.472
 shone-0.433
 tangled-0.431
chy-0.429
 melted-0.429
Token else
Token ID2073
Feature activation+3.47e-01

Pos logit contributions
 how+0.339
ric+0.331
iu+0.323
ext+0.321
 interests+0.320
Neg logit contributions
 faded-0.373
mail-0.350
 healed-0.340
Factory-0.337
 melted-0.325
Token!
Token ID0
Feature activation+5.53e-01

Pos logit contributions
 lighter+0.389
 wax+0.373
 basil+0.368
mail+0.366
Dead+0.365
Neg logit contributions
-0.219
 how-0.213
 as-0.196
ext-0.193
 Just-0.193
Tokenâ
Token ID22940
Feature activation+4.79e-01

Pos logit contributions
 Amen+0.434
ric+0.413
iu+0.411
ext+0.376
 interests+0.346
Neg logit contributions
 faded-0.801
 melted-0.768
 tangled-0.764
 shone-0.752
chy-0.737
Token
Token ID26391
Feature activation+4.95e-01

Pos logit contributions
iu+0.411
 Amen+0.404
ric+0.401
ext+0.366
 interests+0.339
Neg logit contributions
 faded-0.659
 tangled-0.622
 shone-0.618
 melted-0.618
 healed-0.597
Token
Token ID8151
Feature activation+5.11e-01

Pos logit contributions
iu+0.484
 Amen+0.480
ric+0.458
ext+0.436
 interests+0.419
Neg logit contributions
 faded-0.608
 tangled-0.576
 melted-0.573
 shone-0.569
 healed-0.558
Token
Token ID220
Feature activation+5.16e-01

Pos logit contributions
ric+0.592
ext+0.517
iu+0.516
 Amen+0.505
 interests+0.477
Neg logit contributions
 faded-0.568
 shone-0.516
 iv-0.512
 fairy-0.508
 melted-0.507
Token
Token ID198
Feature activation+5.20e-01

Pos logit contributions
ric+0.382
ext+0.333
iu+0.330
 Amen+0.322
 interests+0.290
Neg logit contributions
 faded-0.802
 melted-0.756
 shone-0.745
 tangled-0.733
 disappeared-0.719
Token
Token ID198
Feature activation+5.02e-01

Pos logit contributions
 Amen+0.332
ric+0.310
ext+0.293
iu+0.291
 interests+0.249
Neg logit contributions
 faded-0.853
 melted-0.815
 shone-0.803
 tangled-0.787
chy-0.785
Tokenâ
Token ID22940
Feature activation+4.15e-01

Pos logit contributions
 Amen+0.469
ric+0.438
ext+0.432
iu+0.431
 Wed+0.406
Neg logit contributions
 faded-0.670
 melted-0.639
 shone-0.631
chy-0.621
 vanished-0.617
Token
Token ID26391
Feature activation+4.32e-01

Pos logit contributions
iu+0.231
ric+0.224
 Amen+0.220
ext+0.169
 Wed+0.156
Neg logit contributions
 faded-0.714
 shone-0.678
 tangled-0.676
 melted-0.667
 flowed-0.659
Token
Token ID129
Feature activation+4.37e-01

Pos logit contributions
iu+0.412
ric+0.407
 Amen+0.399
ext+0.376
 interests+0.352
Neg logit contributions
 faded-0.547
 tangled-0.514
 melted-0.511
 shone-0.510
 flowed-0.493
Token
Token ID241
Feature activation+4.33e-01

Pos logit contributions
ric+0.313
iu+0.308
 Amen+0.294
ext+0.277
 Wed+0.251
Neg logit contributions
 faded-0.689
 shone-0.670
 melted-0.647
 tangled-0.636
 disappeared-0.630
Token
Token ID129
Feature activation+6.23e-01

Pos logit contributions
 Amen+0.714
iu+0.704
ric+0.651
ext+0.630
 interests+0.629
Neg logit contributions
 faded-0.742
 tangled-0.701
 melted-0.701
 healed-0.698
 shone-0.687
Token
Token ID241
Feature activation+5.76e-01

Pos logit contributions
iu+0.571
 Amen+0.562
ric+0.551
ext+0.522
 interests+0.511
Neg logit contributions
 faded-0.783
 tangled-0.754
 melted-0.741
 healed-0.729
 shone-0.724
TokenThat
Token ID2504
Feature activation+4.27e-01

Pos logit contributions
iu+0.609
 Amen+0.596
ric+0.579
ext+0.559
 interests+0.524
Neg logit contributions
 faded-0.661
 tangled-0.631
 melted-0.620
 shone-0.617
 healed-0.595
Tokenâ
Token ID22940
Feature activation+4.88e-01

Pos logit contributions
ric+0.380
 Amen+0.351
iu+0.344
 Wed+0.339
 interests+0.332
Neg logit contributions
 faded-0.558
mail-0.524
 tangled-0.512
 melted-0.505
 shone-0.505
Token
Token ID26391
Feature activation+5.02e-01

Pos logit contributions
iu+0.345
ric+0.334
 Amen+0.333
ext+0.288
 interests+0.272
Neg logit contributions
 faded-0.748
 shone-0.704
 tangled-0.703
 melted-0.699
 disappeared-0.682
Token
Token ID8151
Feature activation+5.19e-01

Pos logit contributions
iu+0.447
 Amen+0.441
ric+0.431
ext+0.405
 interests+0.379
Neg logit contributions
 faded-0.662
 tangled-0.627
 melted-0.625
 shone-0.622
 healed-0.606
Tokens
Token ID82
Feature activation+4.46e-01

Pos logit contributions
ric+0.569
ext+0.447
 interests+0.435
 Amen+0.434
!'+0.434
Neg logit contributions
 faded-0.641
 fairy-0.592
 iv-0.576
 shone-0.575
 prick-0.574
Token the
Token ID262
Feature activation+3.84e-01

Pos logit contributions
ric+0.405
iu+0.382
 Amen+0.379
ext+0.348
 interests+0.340
Neg logit contributions
 faded-0.600
 tangled-0.556
 shone-0.556
 melted-0.551
mail-0.542
Token gas
Token ID3623
Feature activation+3.44e-01

Pos logit contributions
ric+0.460
iu+0.441
 Amen+0.441
 Wed+0.412
 interests+0.401
Neg logit contributions
 faded-0.402
mail-0.369
 tangled-0.355
 shone-0.353
 flowed-0.348
Token from
Token ID422
Feature activation+3.87e-01

Pos logit contributions
ric+0.362
iu+0.303
 Amen+0.291
 how+0.291
 interests+0.286
Neg logit contributions
 faded-0.423
 tangled-0.407
mail-0.400
Spider-0.386
Lost-0.383
Token the
Token ID262
Feature activation+3.37e-01

Pos logit contributions
 how+0.285
ric+0.259
 Wed+0.253
 why+0.251
iu+0.250
Neg logit contributions
 faded-0.569
 shone-0.552
mail-0.548
 flowed-0.541
Factory-0.540
Token
Token ID129
Feature activation+6.23e-01

Pos logit contributions
 Amen+0.714
iu+0.704
ric+0.651
ext+0.630
 interests+0.629
Neg logit contributions
 faded-0.742
 tangled-0.701
 melted-0.701
 healed-0.698
 shone-0.687
Token
Token ID241
Feature activation+5.76e-01

Pos logit contributions
iu+0.571
 Amen+0.562
ric+0.551
ext+0.522
 interests+0.511
Neg logit contributions
 faded-0.783
 tangled-0.754
 melted-0.741
 healed-0.729
 shone-0.724
TokenThat
Token ID2504
Feature activation+4.27e-01

Pos logit contributions
iu+0.609
 Amen+0.596
ric+0.579
ext+0.559
 interests+0.524
Neg logit contributions
 faded-0.661
 tangled-0.631
 melted-0.620
 shone-0.617
 healed-0.595
Tokenâ
Token ID22940
Feature activation+4.88e-01

Pos logit contributions
ric+0.380
 Amen+0.351
iu+0.344
 Wed+0.339
 interests+0.332
Neg logit contributions
 faded-0.558
mail-0.524
 tangled-0.512
 melted-0.505
 shone-0.505
Token
Token ID26391
Feature activation+5.02e-01

Pos logit contributions
iu+0.345
ric+0.334
 Amen+0.333
ext+0.288
 interests+0.272
Neg logit contributions
 faded-0.748
 shone-0.704
 tangled-0.703
 melted-0.699
 disappeared-0.682
Token
Token ID8151
Feature activation+5.19e-01

Pos logit contributions
iu+0.447
 Amen+0.441
ric+0.431
ext+0.405
 interests+0.379
Neg logit contributions
 faded-0.662
 tangled-0.627
 melted-0.625
 shone-0.622
 healed-0.606
Tokens
Token ID82
Feature activation+4.46e-01

Pos logit contributions
ric+0.569
ext+0.447
 interests+0.435
 Amen+0.434
!'+0.434
Neg logit contributions
 faded-0.641
 fairy-0.592
 iv-0.576
 shone-0.575
 prick-0.574
Token right
Token ID826
Feature activation+3.62e-01

Pos logit contributions
ric+0.405
iu+0.382
 Amen+0.379
ext+0.348
 interests+0.340
Neg logit contributions
 faded-0.600
 tangled-0.556
 shone-0.556
 melted-0.551
mail-0.542
Token,
Token ID11
Feature activation+4.39e-01

Pos logit contributions
ric+0.252
iu+0.235
 Amen+0.235
 how+0.207
ext+0.202
Neg logit contributions
 faded-0.547
 tangled-0.515
 melted-0.514
mail-0.511
 shone-0.509
Token Emma
Token ID18966
Feature activation+3.26e-01

Pos logit contributions
iu+0.305
ric+0.301
 Amen+0.289
 how+0.271
 interests+0.261
Neg logit contributions
 faded-0.644
 shone-0.608
mail-0.608
 tangled-0.607
 melted-0.598
Token.
Token ID13
Feature activation+4.43e-01

Pos logit contributions
!'+0.215
ric+0.215
 how+0.208
 Amen+0.204
 why+0.189
Neg logit contributions
 faded-0.475
mail-0.462
 darker-0.456
 tangled-0.450
Lost-0.450
Token by
Token ID416
Feature activation+4.68e-01

Pos logit contributions
 faded+0.103
 tangled+0.099
 shone+0.098
 melted+0.096
 whispered+0.095
Neg logit contributions
iu-0.137
ric-0.129
 Amen-0.121
icol-0.120
ext-0.113
Token one
Token ID530
Feature activation+4.35e-01

Pos logit contributions
iu+0.444
 Amen+0.438
ric+0.435
ext+0.406
 interests+0.383
Neg logit contributions
 faded-0.592
 tangled-0.558
 melted-0.555
 shone-0.554
 healed-0.536
Token
Token ID6184
Feature activation+4.61e-01

Pos logit contributions
ric+0.363
 Amen+0.360
 how+0.357
iu+0.338
 interests+0.336
Neg logit contributions
mail-0.558
 faded-0.555
 tangled-0.550
 flowed-0.545
Factory-0.539
Token
Token ID95
Feature activation+4.32e-01

Pos logit contributions
 Amen+0.465
iu+0.464
ric+0.458
ext+0.418
 Wed+0.404
Neg logit contributions
 faded-0.567
 melted-0.543
 tangled-0.540
 shone-0.538
 healed-0.519
Token
Token ID26391
Feature activation+4.75e-01

Pos logit contributions
 Amen+0.371
iu+0.365
ric+0.361
ext+0.325
 Wed+0.320
Neg logit contributions
 faded-0.595
 melted-0.568
 tangled-0.564
 shone-0.562
 disappeared-0.540
Token
Token ID447
Feature activation+5.19e-01

Pos logit contributions
iu+0.487
 Amen+0.478
ric+0.457
ext+0.436
 interests+0.421
Neg logit contributions
 faded-0.557
 shone-0.525
 tangled-0.521
 melted-0.517
 healed-0.515
Token
Token ID250
Feature activation+4.14e-01

Pos logit contributions
 Amen+0.265
ric+0.239
iu+0.175
ext+0.154
icol+0.154
Neg logit contributions
 faded-0.932
 prick-0.899
 iv-0.894
chy-0.893
 leaked-0.892
Token 1
Token ID352
Feature activation+3.12e-01

Pos logit contributions
 Amen+0.427
ric+0.419
iu+0.413
 interests+0.380
 Wed+0.374
Neg logit contributions
 faded-0.494
 melted-0.473
 shone-0.467
 vanished-0.465
 disappeared-0.464
Token,
Token ID11
Feature activation+4.09e-01

Pos logit contributions
iu+0.217
 Amen+0.216
ric+0.215
 how+0.196
 interests+0.184
Neg logit contributions
 faded-0.469
 tangled-0.450
 shone-0.442
 flowed-0.440
chy-0.438
Token 2
Token ID362
Feature activation+2.77e-01

Pos logit contributions
iu+0.305
 Amen+0.300
ric+0.297
 interests+0.260
 how+0.258
Neg logit contributions
 faded-0.594
 tangled-0.568
 shone-0.564
 melted-0.559
 healed-0.548
Token,
Token ID11
Feature activation+3.86e-01

Pos logit contributions
iu+0.186
ric+0.175
 how+0.168
 Amen+0.168
 interests+0.148
Neg logit contributions
 faded-0.432
 tangled-0.412
chy-0.406
 shone-0.405
 flowed-0.402
Tokenâ
Token ID22940
Feature activation+6.14e-01

Pos logit contributions
 faded+0.140
 tangled+0.127
 melted+0.126
 shone+0.124
 healed+0.118
Neg logit contributions
iu-0.288
ric-0.287
 Amen-0.284
ext-0.271
 interests-0.261
Token
Token ID26391
Feature activation+5.98e-01

Pos logit contributions
iu+0.624
 Amen+0.622
ric+0.593
ext+0.584
 interests+0.580
Neg logit contributions
 faded-0.702
 melted-0.677
 tangled-0.673
 healed-0.656
chy-0.654
Token
Token ID220
Feature activation+5.77e-01

Pos logit contributions
iu+0.588
 Amen+0.581
ric+0.562
ext+0.534
 interests+0.512
Neg logit contributions
 faded-0.728
 tangled-0.686
 melted-0.684
 shone-0.680
 healed-0.666
Token
Token ID198
Feature activation+5.56e-01

Pos logit contributions
ric+0.422
 Amen+0.367
iu+0.363
ext+0.361
 Wed+0.315
Neg logit contributions
 faded-0.886
 melted-0.844
 shone-0.831
 tangled-0.824
 flowed-0.797
Token
Token ID220
Feature activation+5.13e-01

Pos logit contributions
 Amen+0.411
ric+0.398
iu+0.388
ext+0.373
 interests+0.342
Neg logit contributions
 faded-0.844
 melted-0.811
 shone-0.799
 tangled-0.780
chy-0.771
Token
Token ID198
Feature activation+5.15e-01

Pos logit contributions
ric+0.373
iu+0.322
 Amen+0.319
ext+0.318
 interests+0.284
Neg logit contributions
 faded-0.804
 melted-0.767
 shone-0.754
 tangled-0.736
 disappeared-0.721
TokenThe
Token ID464
Feature activation+3.85e-01

Pos logit contributions
 Amen+0.493
ric+0.474
iu+0.471
ext+0.456
 interests+0.434
Neg logit contributions
 faded-0.675
 melted-0.644
 shone-0.630
 tangled-0.610
chy-0.608
Token small
Token ID1402
Feature activation+3.34e-01

Pos logit contributions
 Amen+0.438
iu+0.424
ric+0.406
 interests+0.372
 you+0.364
Neg logit contributions
 faded-0.405
mail-0.388
 tangled-0.377
 shone-0.374
 flowed-0.372
Token child
Token ID1200
Feature activation+3.13e-01

Pos logit contributions
 Amen+0.336
iu+0.331
 interests+0.322
ric+0.315
 how+0.307
Neg logit contributions
 faded-0.367
 healed-0.354
mail-0.352
chy-0.350
 flowed-0.340
Token thanked
Token ID26280
Feature activation+3.54e-01

Pos logit contributions
 how+0.279
+0.262
 who+0.251
 why+0.246
 then+0.241
Neg logit contributions
mail-0.411
Lost-0.400
 tangled-0.398
 darker-0.391
 stiff-0.388
Token the
Token ID262
Feature activation+3.17e-01

Pos logit contributions
iu+0.355
 Amen+0.347
ric+0.347
ext+0.324
 interests+0.303
Neg logit contributions
 faded-0.433
 tangled-0.408
 shone-0.406
 melted-0.406
 healed-0.392
Token
Token ID129
Feature activation+6.23e-01

Pos logit contributions
 Amen+0.714
iu+0.704
ric+0.651
ext+0.630
 interests+0.629
Neg logit contributions
 faded-0.742
 tangled-0.701
 melted-0.701
 healed-0.698
 shone-0.687
Token
Token ID241
Feature activation+5.76e-01

Pos logit contributions
iu+0.571
 Amen+0.562
ric+0.551
ext+0.522
 interests+0.511
Neg logit contributions
 faded-0.783
 tangled-0.754
 melted-0.741
 healed-0.729
 shone-0.724
TokenLet
Token ID5756
Feature activation+4.08e-01

Pos logit contributions
iu+0.609
 Amen+0.596
ric+0.579
ext+0.559
 interests+0.524
Neg logit contributions
 faded-0.661
 tangled-0.631
 melted-0.620
 shone-0.617
 healed-0.595
Tokenâ
Token ID22940
Feature activation+4.81e-01

Pos logit contributions
 Amen+0.362
ric+0.360
iu+0.358
ext+0.319
icol+0.287
Neg logit contributions
 faded-0.555
 melted-0.536
 tangled-0.528
 shone-0.521
 healed-0.520
Token
Token ID26391
Feature activation+4.98e-01

Pos logit contributions
iu+0.377
 Amen+0.370
ric+0.361
ext+0.346
 interests+0.316
Neg logit contributions
 faded-0.680
 tangled-0.649
 melted-0.648
 shone-0.642
 healed-0.634
Token
Token ID8151
Feature activation+5.14e-01

Pos logit contributions
iu+0.488
 Amen+0.481
ric+0.443
ext+0.435
 interests+0.413
Neg logit contributions
 faded-0.596
 healed-0.573
 tangled-0.571
 melted-0.564
 shone-0.563
Tokens
Token ID82
Feature activation+4.41e-01

Pos logit contributions
ric+0.590
 Amen+0.545
iu+0.544
ext+0.528
 interests+0.504
Neg logit contributions
 faded-0.566
 melted-0.519
 shone-0.517
 tangled-0.513
 flowed-0.496
Token play
Token ID711
Feature activation+3.56e-01

Pos logit contributions
iu+0.412
 Amen+0.400
ric+0.372
ext+0.368
icol+0.329
Neg logit contributions
 faded-0.578
 tangled-0.552
 melted-0.551
 shone-0.548
 healed-0.540
Token a
Token ID257
Feature activation+3.51e-01

Pos logit contributions
 Amen+0.313
iu+0.309
ric+0.305
ext+0.295
 interests+0.276
Neg logit contributions
 faded-0.473
 melted-0.437
 tangled-0.436
 shone-0.433
chy-0.423
Token joke
Token ID9707
Feature activation+3.15e-01

Pos logit contributions
 Amen+0.388
iu+0.378
ric+0.376
ext+0.354
 interests+0.342
Neg logit contributions
 faded-0.378
 tangled-0.352
 shone-0.344
 flowed-0.343
mail-0.342
Token.
Token ID13
Feature activation+4.41e-01

Pos logit contributions
 how+0.172
 Amen+0.151
ext+0.144
iu+0.142
!'+0.138
Neg logit contributions
 faded-0.520
 tangled-0.502
 melted-0.493
 stiff-0.486
mail-0.484
Token
Token ID129
Feature activation+6.23e-01

Pos logit contributions
 Amen+0.714
iu+0.704
ric+0.651
ext+0.630
 interests+0.629
Neg logit contributions
 faded-0.742
 tangled-0.701
 melted-0.701
 healed-0.698
 shone-0.687
Token
Token ID241
Feature activation+5.76e-01

Pos logit contributions
iu+0.571
 Amen+0.562
ric+0.551
ext+0.522
 interests+0.511
Neg logit contributions
 faded-0.783
 tangled-0.754
 melted-0.741
 healed-0.729
 shone-0.724
TokenLet
Token ID5756
Feature activation+4.08e-01

Pos logit contributions
iu+0.609
 Amen+0.596
ric+0.579
ext+0.559
 interests+0.524
Neg logit contributions
 faded-0.661
 tangled-0.631
 melted-0.620
 shone-0.617
 healed-0.595
Tokenâ
Token ID22940
Feature activation+4.81e-01

Pos logit contributions
 Amen+0.362
ric+0.360
iu+0.358
ext+0.319
icol+0.287
Neg logit contributions
 faded-0.555
 melted-0.536
 tangled-0.528
 shone-0.521
 healed-0.520
Token
Token ID26391
Feature activation+4.98e-01

Pos logit contributions
iu+0.377
 Amen+0.370
ric+0.361
ext+0.346
 interests+0.316
Neg logit contributions
 faded-0.680
 tangled-0.649
 melted-0.648
 shone-0.642
 healed-0.634
Token
Token ID8151
Feature activation+5.14e-01

Pos logit contributions
iu+0.488
 Amen+0.481
ric+0.443
ext+0.435
 interests+0.413
Neg logit contributions
 faded-0.596
 healed-0.573
 tangled-0.571
 melted-0.564
 shone-0.563
Tokens
Token ID82
Feature activation+4.41e-01

Pos logit contributions
ric+0.590
 Amen+0.545
iu+0.544
ext+0.528
 interests+0.504
Neg logit contributions
 faded-0.566
 melted-0.519
 shone-0.517
 tangled-0.513
 flowed-0.496
Token not
Token ID407
Feature activation+4.16e-01

Pos logit contributions
iu+0.412
 Amen+0.400
ric+0.372
ext+0.368
icol+0.329
Neg logit contributions
 faded-0.578
 tangled-0.552
 melted-0.551
 shone-0.548
 healed-0.540
Token think
Token ID892
Feature activation+3.73e-01

Pos logit contributions
iu+0.368
 Amen+0.362
ric+0.354
ext+0.334
 interests+0.306
Neg logit contributions
 faded-0.556
 tangled-0.527
 melted-0.524
 shone-0.520
 healed-0.508
Token about
Token ID546
Feature activation+3.86e-01

Pos logit contributions
iu+0.244
ric+0.243
 how+0.231
 Amen+0.225
ext+0.223
Neg logit contributions
 faded-0.561
 shone-0.535
 tangled-0.531
 melted-0.530
mail-0.528
Token the
Token ID262
Feature activation+3.31e-01

Pos logit contributions
ric+0.299
iu+0.296
 Amen+0.287
ext+0.262
 interests+0.258
Neg logit contributions
 faded-0.556
 tangled-0.524
 melted-0.522
 shone-0.522
 healed-0.509
Token
Token ID129
Feature activation+6.23e-01

Pos logit contributions
 Amen+0.714
iu+0.704
ric+0.651
ext+0.630
 interests+0.629
Neg logit contributions
 faded-0.742
 tangled-0.701
 melted-0.701
 healed-0.698
 shone-0.687
Token
Token ID241
Feature activation+5.76e-01

Pos logit contributions
iu+0.571
 Amen+0.562
ric+0.551
ext+0.522
 interests+0.511
Neg logit contributions
 faded-0.783
 tangled-0.754
 melted-0.741
 healed-0.729
 shone-0.724
TokenLet
Token ID5756
Feature activation+4.08e-01

Pos logit contributions
iu+0.609
 Amen+0.596
ric+0.579
ext+0.559
 interests+0.524
Neg logit contributions
 faded-0.661
 tangled-0.631
 melted-0.620
 shone-0.617
 healed-0.595
Tokenâ
Token ID22940
Feature activation+4.81e-01

Pos logit contributions
 Amen+0.362
ric+0.360
iu+0.358
ext+0.319
icol+0.287
Neg logit contributions
 faded-0.555
 melted-0.536
 tangled-0.528
 shone-0.521
 healed-0.520
Token
Token ID26391
Feature activation+4.98e-01

Pos logit contributions
iu+0.377
 Amen+0.370
ric+0.361
ext+0.346
 interests+0.316
Neg logit contributions
 faded-0.680
 tangled-0.649
 melted-0.648
 shone-0.642
 healed-0.634
Token
Token ID8151
Feature activation+5.14e-01

Pos logit contributions
iu+0.488
 Amen+0.481
ric+0.443
ext+0.435
 interests+0.413
Neg logit contributions
 faded-0.596
 healed-0.573
 tangled-0.571
 melted-0.564
 shone-0.563
Tokens
Token ID82
Feature activation+4.41e-01

Pos logit contributions
ric+0.590
 Amen+0.545
iu+0.544
ext+0.528
 interests+0.504
Neg logit contributions
 faded-0.566
 melted-0.519
 shone-0.517
 tangled-0.513
 flowed-0.496
Token have
Token ID423
Feature activation+3.91e-01

Pos logit contributions
iu+0.412
 Amen+0.400
ric+0.372
ext+0.368
icol+0.329
Neg logit contributions
 faded-0.578
 tangled-0.552
 melted-0.551
 shone-0.548
 healed-0.540
Token a
Token ID257
Feature activation+3.54e-01

Pos logit contributions
iu+0.331
 Amen+0.328
ric+0.315
 interests+0.302
ext+0.301
Neg logit contributions
 faded-0.519
 shone-0.489
 tangled-0.485
chy-0.482
 flowed-0.480
Token stretching
Token ID20880
Feature activation+3.54e-01

Pos logit contributions
 Amen+0.388
iu+0.358
ric+0.353
 interests+0.342
icol+0.331
Neg logit contributions
 faded-0.384
 flowed-0.365
mail-0.360
 tangled-0.360
 iv-0.351
Token race
Token ID3234
Feature activation+3.07e-01

Pos logit contributions
iu+0.378
ric+0.377
 Amen+0.373
ext+0.353
 interests+0.338
Neg logit contributions
 faded-0.397
 tangled-0.373
 melted-0.368
 shone-0.364
 healed-0.355
Token
Token ID129
Feature activation+6.23e-01

Pos logit contributions
 Amen+0.714
iu+0.704
ric+0.651
ext+0.630
 interests+0.629
Neg logit contributions
 faded-0.742
 tangled-0.701
 melted-0.701
 healed-0.698
 shone-0.687
Token
Token ID241
Feature activation+5.76e-01

Pos logit contributions
iu+0.571
 Amen+0.562
ric+0.551
ext+0.522
 interests+0.511
Neg logit contributions
 faded-0.783
 tangled-0.754
 melted-0.741
 healed-0.729
 shone-0.724
TokenLet
Token ID5756
Feature activation+4.08e-01

Pos logit contributions
iu+0.609
 Amen+0.596
ric+0.579
ext+0.559
 interests+0.524
Neg logit contributions
 faded-0.661
 tangled-0.631
 melted-0.620
 shone-0.617
 healed-0.595
Tokenâ
Token ID22940
Feature activation+4.81e-01

Pos logit contributions
 Amen+0.362
ric+0.360
iu+0.358
ext+0.319
icol+0.287
Neg logit contributions
 faded-0.555
 melted-0.536
 tangled-0.528
 shone-0.521
 healed-0.520
Token
Token ID26391
Feature activation+4.98e-01

Pos logit contributions
iu+0.377
 Amen+0.370
ric+0.361
ext+0.346
 interests+0.316
Neg logit contributions
 faded-0.680
 tangled-0.649
 melted-0.648
 shone-0.642
 healed-0.634
Token
Token ID8151
Feature activation+5.14e-01

Pos logit contributions
iu+0.488
 Amen+0.481
ric+0.443
ext+0.435
 interests+0.413
Neg logit contributions
 faded-0.596
 healed-0.573
 tangled-0.571
 melted-0.564
 shone-0.563
Tokens
Token ID82
Feature activation+4.41e-01

Pos logit contributions
ric+0.590
 Amen+0.545
iu+0.544
ext+0.528
 interests+0.504
Neg logit contributions
 faded-0.566
 melted-0.519
 shone-0.517
 tangled-0.513
 flowed-0.496
Token explore
Token ID7301
Feature activation+3.46e-01

Pos logit contributions
iu+0.412
 Amen+0.400
ric+0.372
ext+0.368
icol+0.329
Neg logit contributions
 faded-0.578
 tangled-0.552
 melted-0.551
 shone-0.548
 healed-0.540
Token and
Token ID290
Feature activation+4.05e-01

Pos logit contributions
ric+0.301
iu+0.296
 Amen+0.295
ext+0.281
 how+0.277
Neg logit contributions
 faded-0.454
 tangled-0.420
 melted-0.420
 shone-0.415
chy-0.411
Token have
Token ID423
Feature activation+3.44e-01

Pos logit contributions
iu+0.406
 Amen+0.400
ric+0.392
ext+0.373
 interests+0.342
Neg logit contributions
 faded-0.499
 tangled-0.473
 melted-0.469
 shone-0.469
 healed-0.452
Token some
Token ID617
Feature activation+2.85e-01

Pos logit contributions
iu+0.308
 Amen+0.305
ric+0.297
 interests+0.295
ext+0.289
Neg logit contributions
 faded-0.432
 tangled-0.403
 shone-0.403
chy-0.400
 melted-0.397
Token
Token ID129
Feature activation+6.23e-01

Pos logit contributions
 Amen+0.714
iu+0.704
ric+0.651
ext+0.630
 interests+0.629
Neg logit contributions
 faded-0.742
 tangled-0.701
 melted-0.701
 healed-0.698
 shone-0.687
Token
Token ID241
Feature activation+5.76e-01

Pos logit contributions
iu+0.571
 Amen+0.562
ric+0.551
ext+0.522
 interests+0.511
Neg logit contributions
 faded-0.783
 tangled-0.754
 melted-0.741
 healed-0.729
 shone-0.724
TokenWho
Token ID8241
Feature activation+4.24e-01

Pos logit contributions
iu+0.609
 Amen+0.596
ric+0.579
ext+0.559
 interests+0.524
Neg logit contributions
 faded-0.661
 tangled-0.631
 melted-0.620
 shone-0.617
 healed-0.595
Tokenâ
Token ID22940
Feature activation+4.76e-01

Pos logit contributions
 interests+0.371
ric+0.366
iu+0.363
 how+0.362
 Amen+0.357
Neg logit contributions
 faded-0.535
 tangled-0.510
 melted-0.491
 stiff-0.489
 shone-0.487
Token
Token ID26391
Feature activation+4.94e-01

Pos logit contributions
iu+0.370
 Amen+0.363
ric+0.358
ext+0.329
 interests+0.305
Neg logit contributions
 faded-0.686
 tangled-0.651
 melted-0.649
 shone-0.647
 healed-0.629
Token
Token ID8151
Feature activation+5.12e-01

Pos logit contributions
iu+0.437
 Amen+0.429
ric+0.421
ext+0.394
 interests+0.370
Neg logit contributions
 faded-0.656
 tangled-0.620
 melted-0.618
 shone-0.616
 healed-0.601
Tokens
Token ID82
Feature activation+4.40e-01

Pos logit contributions
ric+0.606
 Amen+0.509
ext+0.497
iu+0.494
 how+0.484
Neg logit contributions
 faded-0.578
 melted-0.525
 shone-0.520
 tangled-0.513
mail-0.504
Token there
Token ID612
Feature activation+3.93e-01

Pos logit contributions
ric+0.281
 Amen+0.267
iu+0.261
ext+0.233
 interests+0.218
Neg logit contributions
 faded-0.695
 tangled-0.666
 melted-0.658
 shone-0.655
mail-0.639
Token?
Token ID30
Feature activation+4.42e-01

Pos logit contributions
ric+0.188
 how+0.142
 why+0.136
 you+0.135
 Wed+0.124
Neg logit contributions
 faded-0.714
 tangled-0.677
mail-0.663
 melted-0.656
 stiff-0.655
Tokenâ
Token ID22940
Feature activation+3.90e-01

Pos logit contributions
 Amen+0.301
ric+0.295
iu+0.267
 We+0.240
 Wed+0.237
Neg logit contributions
 faded-0.703
 melted-0.681
 shone-0.655
 tangled-0.653
chy-0.648
Token
Token ID26391
Feature activation+4.11e-01

Pos logit contributions
ric+0.235
iu+0.229
 Amen+0.227
ext+0.163
 Wed+0.162
Neg logit contributions
 faded-0.662
 tangled-0.618
 shone-0.616
 melted-0.612
 disappeared-0.604
Token
Token ID241
Feature activation+5.96e-01

Pos logit contributions
 interests+0.680
 Amen+0.672
iu+0.644
ric+0.627
pieces+0.576
Neg logit contributions
 faded-0.711
 tangled-0.710
 healed-0.704
 melted-0.689
chy-0.686
TokenPlease
Token ID5492
Feature activation+4.89e-01

Pos logit contributions
iu+0.688
 Amen+0.657
ric+0.620
ext+0.620
icol+0.618
Neg logit contributions
 faded-0.616
 tangled-0.610
 shone-0.583
 melted-0.581
 flowed-0.559
Token don
Token ID836
Feature activation+4.57e-01

Pos logit contributions
ric+0.321
 Amen+0.316
iu+0.308
 interests+0.274
 how+0.266
Neg logit contributions
 faded-0.760
 shone-0.719
 tangled-0.718
 flowed-0.705
 melted-0.704
Tokenâ
Token ID22940
Feature activation+4.77e-01

Pos logit contributions
iu+0.677
ric+0.669
icol+0.658
 interests+0.658
 Amen+0.600
Neg logit contributions
 tangled-0.370
 healed-0.348
 hurt-0.344
 faded-0.340
 weak-0.331
Token
Token ID26391
Feature activation+4.91e-01

Pos logit contributions
iu+0.414
ric+0.407
 Amen+0.395
ext+0.386
 interests+0.360
Neg logit contributions
 faded-0.624
 melted-0.600
 tangled-0.597
 healed-0.591
chy-0.580
Token
Token ID8151
Feature activation+5.07e-01

Pos logit contributions
iu+0.485
 Amen+0.471
ric+0.458
ext+0.436
 interests+0.412
Neg logit contributions
 faded-0.591
 tangled-0.563
 healed-0.563
 melted-0.558
 shone-0.549
Tokent
Token ID83
Feature activation+3.95e-01

Pos logit contributions
ric+0.573
 Amen+0.540
iu+0.535
ext+0.524
 interests+0.494
Neg logit contributions
 faded-0.559
 shone-0.522
 melted-0.512
 tangled-0.509
 flowed-0.506
Token eat
Token ID4483
Feature activation+3.63e-01

Pos logit contributions
ric+0.291
 Amen+0.290
iu+0.281
 interests+0.255
ext+0.248
Neg logit contributions
 faded-0.580
 shone-0.556
 tangled-0.548
 melted-0.546
 flowed-0.538
Token me
Token ID502
Feature activation+3.65e-01

Pos logit contributions
ric+0.252
iu+0.247
 Amen+0.232
 how+0.231
ext+0.223
Neg logit contributions
 faded-0.549
 shone-0.515
 tangled-0.504
 flowed-0.500
 melted-0.500
Token,
Token ID11
Feature activation+4.20e-01

Pos logit contributions
ric+0.209
 how+0.193
 interests+0.188
 Amen+0.184
iu+0.183
Neg logit contributions
 faded-0.599
 flowed-0.565
 shone-0.557
 tangled-0.557
mail-0.551
Tokenâ
Token ID22940
Feature activation+3.90e-01

Pos logit contributions
iu+0.283
 Amen+0.279
ric+0.276
ext+0.245
 interests+0.240
Neg logit contributions
 faded-0.630
 shone-0.599
 tangled-0.593
 melted-0.588
mail-0.587
Token ash
Token ID12530
Feature activation+4.26e-01

Pos logit contributions
 faded+0.142
 shone+0.133
 tangled+0.131
 melted+0.130
chy+0.129
Neg logit contributions
iu-0.139
ric-0.134
 Amen-0.128
ext-0.123
icol-0.112
Tokentr
Token ID2213
Feature activation+5.27e-01

Pos logit contributions
iu+0.430
ric+0.425
 Amen+0.412
ext+0.391
 interests+0.361
Neg logit contributions
 faded-0.502
 tangled-0.473
 shone-0.456
 melted-0.455
mail-0.448
Tokenay
Token ID323
Feature activation+4.74e-01

Pos logit contributions
 Amen+0.624
iu+0.618
 interests+0.591
ric+0.580
ext+0.563
Neg logit contributions
 faded-0.562
 melted-0.551
 shone-0.550
 tangled-0.543
 disappeared-0.536
Token!
Token ID0
Feature activation+5.15e-01

Pos logit contributions
ric+0.407
 Amen+0.407
iu+0.402
 how+0.357
ext+0.351
Neg logit contributions
 faded-0.631
 tangled-0.597
 stiff-0.582
 flowed-0.571
chy-0.561
Token
Token ID198
Feature activation+5.22e-01

Pos logit contributions
 Amen+0.416
iu+0.383
ric+0.370
ext+0.340
 interests+0.319
Neg logit contributions
 faded-0.749
 melted-0.721
 tangled-0.715
 shone-0.701
chy-0.697
Token
Token ID198
Feature activation+5.06e-01

Pos logit contributions
 Amen+0.435
iu+0.423
ric+0.419
ext+0.394
 interests+0.365
Neg logit contributions
 faded-0.734
 melted-0.699
 shone-0.693
 tangled-0.691
 flowed-0.669
TokenHe
Token ID1544
Feature activation+3.98e-01

Pos logit contributions
 Amen+0.560
iu+0.557
ric+0.544
ext+0.521
 interests+0.495
Neg logit contributions
 faded-0.568
 melted-0.533
 shone-0.529
 tangled-0.528
 healed-0.507
Token brought
Token ID3181
Feature activation+3.06e-01

Pos logit contributions
 how+0.438
 Amen+0.437
iu+0.430
ext+0.407
ric+0.393
Neg logit contributions
 tangled-0.391
mail-0.386
Factory-0.385
 faded-0.384
Ancient-0.382
Token it
Token ID340
Feature activation+2.90e-01

Pos logit contributions
iu+0.187
 how+0.173
 Amen+0.160
ric+0.157
ext+0.146
Neg logit contributions
 faded-0.504
chy-0.472
 flowed-0.461
 shone-0.457
 tangled-0.454
Token inside
Token ID2641
Feature activation+2.89e-01

Pos logit contributions
 how+0.140
 where+0.126
 what+0.106
 as+0.093
 why+0.083
Neg logit contributions
 faded-0.519
Dead-0.500
Ancient-0.497
 flowed-0.496
opy-0.494
Token,
Token ID11
Feature activation+3.83e-01

Pos logit contributions
 how+0.178
 where+0.156
 what+0.143
 Amen+0.142
 why+0.139
Neg logit contributions
 faded-0.472
 tangled-0.452
chy-0.446
 flowed-0.446
Factory-0.439
Token here
Token ID994
Feature activation+4.69e-01

Pos logit contributions
 whispered+0.084
 spoke+0.083
 faded+0.081
mail+0.078
Ancient+0.077
Neg logit contributions
ric-0.187
 Amen-0.177
iu-0.172
ext-0.160
 Wed-0.154
Token?
Token ID30
Feature activation+5.10e-01

Pos logit contributions
iu+0.359
 Amen+0.349
ric+0.344
ext+0.316
 interests+0.291
Neg logit contributions
 faded-0.682
 melted-0.646
 tangled-0.646
 shone-0.644
 healed-0.627
Tokenâ
Token ID22940
Feature activation+4.55e-01

Pos logit contributions
 Amen+0.296
ric+0.258
iu+0.247
ext+0.214
 interests+0.207
Neg logit contributions
 faded-0.866
 melted-0.840
 tangled-0.827
 shone-0.816
chy-0.809
Token
Token ID26391
Feature activation+4.74e-01

Pos logit contributions
 Amen+0.352
iu+0.341
ric+0.338
ext+0.284
 interests+0.260
Neg logit contributions
 faded-0.687
 tangled-0.650
 shone-0.646
 melted-0.640
 flowed-0.625
Token
Token ID220
Feature activation+4.95e-01

Pos logit contributions
iu+0.476
 Amen+0.463
ric+0.439
ext+0.425
 interests+0.410
Neg logit contributions
 faded-0.568
 tangled-0.540
 melted-0.534
 shone-0.532
 healed-0.530
Token
Token ID198
Feature activation+5.06e-01

Pos logit contributions
ric+0.368
 Amen+0.339
iu+0.331
ext+0.320
 interests+0.286
Neg logit contributions
 faded-0.758
 melted-0.717
 shone-0.706
 tangled-0.698
 disappeared-0.682
Token
Token ID198
Feature activation+4.90e-01

Pos logit contributions
 Amen+0.298
ric+0.274
ext+0.252
iu+0.251
 Wed+0.219
Neg logit contributions
 faded-0.861
 melted-0.832
 shone-0.822
 vanished-0.801
chy-0.800
TokenThe
Token ID464
Feature activation+3.61e-01

Pos logit contributions
 Amen+0.460
ric+0.429
iu+0.427
ext+0.421
 Wed+0.402
Neg logit contributions
 faded-0.651
 melted-0.626
 shone-0.624
chy-0.603
 vanished-0.601
Token box
Token ID3091
Feature activation+3.07e-01

Pos logit contributions
 Amen+0.427
iu+0.399
ric+0.376
 interests+0.355
 you+0.350
Neg logit contributions
 faded-0.367
mail-0.359
 tangled-0.347
 flowed-0.345
 shone-0.340
Token answered
Token ID9373
Feature activation+3.09e-01

Pos logit contributions
 how+0.245
 that+0.223
ext+0.219
 we+0.218
iu+0.213
Neg logit contributions
Lost-0.376
 Berry-0.375
Ancient-0.374
Factory-0.372
mail-0.360
Token,
Token ID11
Feature activation+3.81e-01

Pos logit contributions
 how+0.215
 why+0.178
 what+0.168
 where+0.156
ap+0.142
Neg logit contributions
 tangled-0.515
 faded-0.484
Factory-0.484
 melted-0.483
mail-0.481
Token always
Token ID1464
Feature activation-2.26e-01

Pos logit contributions
 faded+0.261
 melted+0.245
 tangled+0.236
chy+0.235
 healed+0.234
Neg logit contributions
 Amen-0.213
iu-0.211
ric-0.209
icol-0.196
ext-0.194
Token make
Token ID787
Feature activation-2.59e-01

Pos logit contributions
 faded+0.272
 healed+0.252
 melted+0.251
 dead+0.248
 tangled+0.244
Neg logit contributions
iu-0.246
 Amen-0.240
ext-0.238
ric-0.236
icol-0.226
Token things
Token ID1243
Feature activation-2.66e-01

Pos logit contributions
 faded+0.340
 tangled+0.324
 melted+0.322
 shone+0.315
 healed+0.313
Neg logit contributions
 Amen-0.235
iu-0.235
ric-0.227
ext-0.208
 interests-0.193
Token work
Token ID670
Feature activation-2.78e-01

Pos logit contributions
 faded+0.263
 tangled+0.251
 melted+0.246
 healed+0.240
chy+0.235
Neg logit contributions
iu-0.345
 Amen-0.341
ric-0.330
ext-0.318
icol-0.298
Token out
Token ID503
Feature activation-2.72e-01

Pos logit contributions
 faded+0.369
 shone+0.355
 tangled+0.347
mail+0.346
 melted+0.346
Neg logit contributions
iu-0.226
ric-0.217
 Amen-0.213
ext-0.206
 interests-0.203
Token perfectly
Token ID7138
Feature activation-3.56e-01

Pos logit contributions
 faded+0.365
 tangled+0.347
 melted+0.346
 shone+0.345
 healed+0.334
Neg logit contributions
iu-0.228
ric-0.227
 Amen-0.224
ext-0.207
 interests-0.201
Token.
Token ID13
Feature activation-1.27e-01

Pos logit contributions
 faded+0.529
 shone+0.523
Spider+0.522
Bee+0.519
mail+0.518
Neg logit contributions
 how-0.200
ric-0.199
iu-0.197
 interests-0.195
ext-0.176
Token Sometimes
Token ID8975
Feature activation-2.16e-01

Pos logit contributions
 faded+0.175
 tangled+0.167
 melted+0.167
 shone+0.166
 flowed+0.160
Neg logit contributions
iu-0.106
 Amen-0.105
ric-0.104
ext-0.097
 interests-0.090
Token,
Token ID11
Feature activation-1.72e-01

Pos logit contributions
 faded+0.283
 tangled+0.267
 melted+0.266
 shone+0.265
 healed+0.257
Neg logit contributions
iu-0.197
 Amen-0.194
ric-0.192
ext-0.179
 interests-0.167
Token things
Token ID1243
Feature activation-2.80e-01

Pos logit contributions
 faded+0.227
 tangled+0.214
 melted+0.213
 shone+0.212
 healed+0.206
Neg logit contributions
iu-0.156
 Amen-0.153
ric-0.151
ext-0.141
 interests-0.132
Token just
Token ID655
Feature activation-2.37e-01

Pos logit contributions
 faded+0.351
 shone+0.331
 tangled+0.330
 melted+0.329
mail+0.321
Neg logit contributions
iu-0.251
ric-0.249
 Amen-0.245
ext-0.235
 interests-0.226
Token home
Token ID1363
Feature activation-2.83e-01

Pos logit contributions
 tangled+0.234
 faded+0.229
 dark+0.217
 shone+0.216
 disappeared+0.214
Neg logit contributions
iu-0.317
 Amen-0.304
ric-0.301
ext-0.294
 interests-0.281
Token and
Token ID290
Feature activation-1.95e-01

Pos logit contributions
 faded+0.437
 shone+0.433
 flowed+0.422
 healed+0.417
 tangled+0.416
Neg logit contributions
iu-0.171
 how-0.157
ric-0.151
 Amen-0.146
 where-0.141
Token then
Token ID788
Feature activation-2.51e-01

Pos logit contributions
Bee+0.205
 iv+0.205
 faded+0.204
chy+0.203
 tangled+0.201
Neg logit contributions
 how-0.213
 why-0.200
iu-0.196
 where-0.195
 Amen-0.193
Token they
Token ID484
Feature activation-2.11e-01

Pos logit contributions
Bee+0.370
 iv+0.336
Fox+0.334
Mouse+0.332
rito+0.323
Neg logit contributions
 how-0.236
 why-0.213
 what-0.211
 go-0.208
 spend-0.205
Token sne
Token ID10505
Feature activation-2.71e-01

Pos logit contributions
Bee+0.264
 swollen+0.262
 tangled+0.257
chy+0.254
 iv+0.252
Neg logit contributions
 spend-0.190
 how-0.189
 go-0.183
 why-0.178
iu-0.172
Tokenaked
Token ID4335
Feature activation-3.55e-01

Pos logit contributions
 faded+0.180
 tangled+0.163
chy+0.152
 dead+0.152
 melted+0.151
Neg logit contributions
iu-0.425
 Amen-0.414
ric-0.412
 interests-0.404
ext-0.401
Token in
Token ID287
Feature activation-2.52e-01

Pos logit contributions
 faded+0.436
 shone+0.408
 tangled+0.407
 melted+0.407
 healed+0.395
Neg logit contributions
iu-0.354
 Amen-0.344
ric-0.336
ext-0.321
 interests-0.302
Token.
Token ID13
Feature activation-1.29e-01

Pos logit contributions
 faded+0.301
 tangled+0.283
 melted+0.282
 shone+0.278
 healed+0.268
Neg logit contributions
iu-0.261
ric-0.258
 Amen-0.256
ext-0.241
 interests-0.226
Token
Token ID198
Feature activation-1.14e-01

Pos logit contributions
 faded+0.142
 tangled+0.136
 melted+0.134
 shone+0.132
 flowed+0.131
Neg logit contributions
 Amen-0.146
iu-0.143
ric-0.140
ext-0.128
 interests-0.122
Token
Token ID198
Feature activation-1.15e-01

Pos logit contributions
 faded+0.121
 tangled+0.115
 melted+0.112
mail+0.111
 shone+0.111
Neg logit contributions
iu-0.134
ric-0.131
 Amen-0.126
ext-0.113
icol-0.112
TokenThe
Token ID464
Feature activation-2.45e-01

Pos logit contributions
 faded+0.119
 tangled+0.113
 melted+0.112
 shone+0.108
 flowed+0.106
Neg logit contributions
ric-0.140
iu-0.138
 Amen-0.137
ext-0.124
 interests-0.119
Token,
Token ID11
Feature activation-1.56e-01

Pos logit contributions
 faded+0.424
chy+0.420
 melted+0.419
 healed+0.418
 flowed+0.412
Neg logit contributions
 how-0.149
ap-0.121
 what-0.120
ext-0.119
 interests-0.115
Token the
Token ID262
Feature activation-2.44e-01

Pos logit contributions
chy+0.252
 faded+0.244
mail+0.243
 tangled+0.240
 iv+0.236
Neg logit contributions
 how-0.092
 what-0.077
 we-0.076
iu-0.072
 why-0.071
Token band
Token ID4097
Feature activation-2.85e-01

Pos logit contributions
 faded+0.298
mail+0.295
 hatched+0.288
 shone+0.285
 iv+0.282
Neg logit contributions
 Amen-0.228
ext-0.202
iu-0.198
 interests-0.194
ric-0.189
Token was
Token ID373
Feature activation-2.38e-01

Pos logit contributions
 iv+0.262
 rotten+0.261
 tangled+0.256
mail+0.254
 darker+0.249
Neg logit contributions
iu-0.326
 Amen-0.320
ric-0.316
 interests-0.316
ext-0.309
Token playing
Token ID2712
Feature activation-2.55e-01

Pos logit contributions
Factory+0.337
mail+0.336
Bee+0.325
Ancient+0.317
Supp+0.314
Neg logit contributions
 how-0.228
 why-0.188
 about-0.185
 where-0.184
 what-0.181
Token perfectly
Token ID7138
Feature activation-3.55e-01

Pos logit contributions
 faded+0.387
 melted+0.368
mail+0.368
 shone+0.366
 tangled+0.365
Neg logit contributions
 how-0.175
ap-0.155
 Amen-0.155
ext-0.153
 where-0.140
Token!
Token ID0
Feature activation-1.30e-01

Pos logit contributions
 faded+0.527
mail+0.517
 shone+0.507
 flowed+0.500
Bee+0.499
Neg logit contributions
iu-0.218
 how-0.206
ext-0.199
ric-0.198
 Amen-0.187
Token The
Token ID383
Feature activation-2.25e-01

Pos logit contributions
 faded+0.213
 tangled+0.208
 melted+0.207
 shone+0.205
chy+0.200
Neg logit contributions
 Amen-0.071
iu-0.071
ric-0.070
ext-0.069
 We-0.067
Token animals
Token ID4695
Feature activation-2.71e-01

Pos logit contributions
 faded+0.253
mail+0.248
 hatched+0.245
 leaked+0.238
 shone+0.238
Neg logit contributions
 Amen-0.238
ext-0.210
iu-0.205
ric-0.199
 interests-0.199
Token cheered
Token ID34767
Feature activation-2.55e-01

Pos logit contributions
 stiff+0.428
mail+0.428
chy+0.427
 Swiss+0.425
 bland+0.425
Neg logit contributions
 how-0.209
 who-0.202
 each-0.159
 that-0.156
 what-0.152
Token and
Token ID290
Feature activation-1.99e-01

Pos logit contributions
 faded+0.336
 tangled+0.322
 melted+0.318
 shone+0.318
 healed+0.308
Neg logit contributions
iu-0.227
 Amen-0.223
ric-0.217
ext-0.210
 interests-0.195
Token wanted
Token ID2227
Feature activation-2.29e-01

Pos logit contributions
 swarm+0.262
Ancient+0.242
 iv+0.241
Factory+0.237
 heavens+0.237
Neg logit contributions
iu-0.135
 how-0.135
n-0.132
 just-0.131
 then-0.123
Token something
Token ID1223
Feature activation-2.52e-01

Pos logit contributions
 faded+0.412
Bee+0.398
 shone+0.393
 swarm+0.392
 melted+0.391
Neg logit contributions
 how-0.095
ric-0.072
 interests-0.072
 Wed-0.071
 why-0.059
Token that
Token ID326
Feature activation-2.01e-01

Pos logit contributions
 faded+0.300
mail+0.279
 shone+0.278
 healed+0.275
 melted+0.274
Neg logit contributions
iu-0.237
 how-0.236
ext-0.229
ric-0.229
 interests-0.228
Token fit
Token ID4197
Feature activation-3.01e-01

Pos logit contributions
Bee+0.267
 swarm+0.260
Factory+0.259
 faded+0.254
opy+0.251
Neg logit contributions
 how-0.179
 you-0.164
 why-0.156
 Wed-0.149
 go-0.149
Token just
Token ID655
Feature activation-2.22e-01

Pos logit contributions
 faded+0.452
Factory+0.431
mail+0.429
 swarm+0.429
Bee+0.429
Neg logit contributions
 how-0.195
iu-0.178
ric-0.175
 interests-0.171
ext-0.167
Token perfectly
Token ID7138
Feature activation-3.53e-01

Pos logit contributions
 faded+0.336
Factory+0.336
mail+0.334
Bee+0.333
Spider+0.329
Neg logit contributions
 how-0.155
ric-0.137
 why-0.130
iu-0.128
 what-0.123
Token.
Token ID13
Feature activation-1.13e-01

Pos logit contributions
Spider+0.610
 swarm+0.604
Bee+0.603
 faded+0.594
mail+0.591
Neg logit contributions
 how-0.156
!'-0.125
 where-0.118
iu-0.116
 as-0.111
Token Then
Token ID3244
Feature activation-1.98e-01

Pos logit contributions
 tangled+0.091
mail+0.090
 faded+0.090
 melted+0.089
 whispered+0.088
Neg logit contributions
 Amen-0.160
ric-0.157
iu-0.154
pieces-0.150
ext-0.140
Token Tommy
Token ID19919
Feature activation-2.75e-01

Pos logit contributions
 faded+0.189
 shone+0.174
 melted+0.173
 disappeared+0.171
 vanished+0.166
Neg logit contributions
ric-0.286
iu-0.263
 Amen-0.256
ext-0.245
pieces-0.242
Token went
Token ID1816
Feature activation-2.47e-01

Pos logit contributions
 swarm+0.351
 iv+0.350
iest+0.346
 lighter+0.346
 faded+0.345
Neg logit contributions
-0.220
ast-0.195
 how-0.193
 spend-0.187
iu-0.186
Token to
Token ID284
Feature activation-2.20e-01

Pos logit contributions
 faded+0.338
 tangled+0.325
 shone+0.319
 melted+0.319
 healed+0.309
Neg logit contributions
iu-0.216
 Amen-0.211
ric-0.209
ext-0.192
 interests-0.179
Token "
Token ID366
Feature activation-1.80e-01

Pos logit contributions
 faded+0.122
 tangled+0.114
 melted+0.114
 shone+0.113
 healed+0.109
Neg logit contributions
iu-0.115
 Amen-0.114
ric-0.113
ext-0.106
 interests-0.101
TokenThis
Token ID1212
Feature activation-2.58e-01

Pos logit contributions
 faded+0.143
 shone+0.133
 tangled+0.129
 melted+0.128
 healed+0.124
Neg logit contributions
ric-0.262
iu-0.258
 Amen-0.249
ext-0.244
 interests-0.235
Token swing
Token ID9628
Feature activation-2.70e-01

Pos logit contributions
 faded+0.305
 tangled+0.289
 melted+0.289
 shone+0.286
 healed+0.276
Neg logit contributions
iu-0.273
 Amen-0.273
ric-0.264
ext-0.248
 interests-0.232
Token fits
Token ID11414
Feature activation-3.14e-01

Pos logit contributions
 faded+0.423
Ancient+0.408
mail+0.408
Factory+0.407
Dead+0.405
Neg logit contributions
 you-0.154
 how-0.135
ric-0.129
 we-0.128
 We-0.119
Token me
Token ID502
Feature activation-2.45e-01

Pos logit contributions
 faded+0.564
 flowed+0.527
 shone+0.524
 tangled+0.519
 melted+0.518
Neg logit contributions
iu-0.146
 Amen-0.143
ext-0.123
ric-0.120
 interests-0.111
Token perfectly
Token ID7138
Feature activation-3.52e-01

Pos logit contributions
 faded+0.359
 shone+0.335
 flowed+0.331
 melted+0.329
 tangled+0.328
Neg logit contributions
iu-0.188
ric-0.182
 Amen-0.179
ext-0.171
 interests-0.155
Token,
Token ID11
Feature activation-1.65e-01

Pos logit contributions
 faded+0.575
 shone+0.553
 flowed+0.547
 melted+0.537
 tangled+0.535
Neg logit contributions
iu-0.198
 Amen-0.193
ext-0.166
ric-0.166
 interests-0.150
Token Mom
Token ID11254
Feature activation-2.59e-01

Pos logit contributions
 faded+0.204
 tangled+0.192
 melted+0.192
 shone+0.191
 healed+0.185
Neg logit contributions
iu-0.164
 Amen-0.161
ric-0.160
ext-0.149
 interests-0.140
Token!"
Token ID2474
Feature activation-1.38e-01

Pos logit contributions
 faded+0.303
 tangled+0.295
 melted+0.293
 shone+0.291
 healed+0.287
Neg logit contributions
iu-0.270
ric-0.259
ext-0.258
 Amen-0.257
 interests-0.232
Token Lily
Token ID20037
Feature activation-2.82e-01

Pos logit contributions
 faded+0.115
 tangled+0.106
 shone+0.104
 melted+0.103
 whispered+0.102
Neg logit contributions
ric-0.204
iu-0.198
 Amen-0.196
ext-0.184
 interests-0.178
Token said
Token ID531
Feature activation-2.14e-01

Pos logit contributions
 faded+0.243
 flowed+0.219
 iv+0.218
 tangled+0.214
chy+0.212
Neg logit contributions
iu-0.377
 Amen-0.354
 how-0.351
ext-0.347
ric-0.340
Token the
Token ID262
Feature activation-2.40e-01

Pos logit contributions
 faded+0.362
 flowed+0.329
 shone+0.329
chy+0.328
 iv+0.327
Neg logit contributions
 Wed-0.193
iu-0.192
 how-0.185
ext-0.167
ric-0.167
Token skirt
Token ID23967
Feature activation-2.92e-01

Pos logit contributions
 faded+0.302
 shone+0.275
 melted+0.271
 flowed+0.271
 tangled+0.269
Neg logit contributions
iu-0.224
 Amen-0.211
 interests-0.206
ext-0.203
ric-0.194
Token and
Token ID290
Feature activation-1.83e-01

Pos logit contributions
 swarm+0.500
Bee+0.491
 faded+0.487
 bothered+0.482
Spider+0.477
Neg logit contributions
 how-0.135
 where-0.128
 as-0.100
 that-0.099
ap-0.084
Token it
Token ID340
Feature activation-2.72e-01

Pos logit contributions
chy+0.226
 iv+0.215
 faded+0.211
 stiff+0.206
 tangled+0.206
Neg logit contributions
 how-0.181
iu-0.169
 spend-0.148
 Amen-0.147
 what-0.143
Token fit
Token ID4197
Feature activation-3.00e-01

Pos logit contributions
Bee+0.503
 swarm+0.481
Pr+0.470
Lost+0.469
Mouse+0.463
Neg logit contributions
 how-0.173
 that-0.141
 go-0.126
 just-0.121
 as-0.105
Token perfectly
Token ID7138
Feature activation-3.51e-01

Pos logit contributions
 swarm+0.495
Dead+0.492
Spider+0.474
Factory+0.472
Bee+0.470
Neg logit contributions
 how-0.212
 what-0.163
 where-0.162
 as-0.162
 you-0.141
Token.
Token ID13
Feature activation-1.18e-01

Pos logit contributions
 faded+0.548
Spider+0.545
Bee+0.544
 swarm+0.538
Factory+0.525
Neg logit contributions
iu-0.207
 how-0.195
 where-0.159
!'-0.156
ric-0.156
Token She
Token ID1375
Feature activation-1.99e-01

Pos logit contributions
 faded+0.189
 melted+0.180
 tangled+0.179
 shone+0.178
chy+0.174
Neg logit contributions
iu-0.073
 Amen-0.073
ric-0.067
ext-0.063
 interests-0.062
Token was
Token ID373
Feature activation-2.47e-01

Pos logit contributions
 darker+0.249
 iv+0.241
 swarm+0.240
Bee+0.232
 triggers+0.231
Neg logit contributions
 how-0.196
iu-0.172
 interests-0.157
ext-0.156
 spend-0.153
Token so
Token ID523
Feature activation-1.98e-01

Pos logit contributions
 iv+0.436
Bee+0.433
Factory+0.426
 faded+0.423
mail+0.422
Neg logit contributions
 how-0.142
ric-0.100
 Amen-0.099
iu-0.095
 where-0.091
Token happy
Token ID3772
Feature activation-2.63e-01

Pos logit contributions
 iv+0.338
Factory+0.290
Bee+0.285
 White+0.281
 thorn+0.279
Neg logit contributions
 how-0.146
iu-0.124
 Amen-0.117
ext-0.116
 you-0.114
Token feeling
Token ID4203
Feature activation-2.72e-01

Pos logit contributions
 faded+0.335
mail+0.313
 tangled+0.307
 melted+0.307
 iv+0.306
Neg logit contributions
 Amen-0.208
ric-0.205
iu-0.199
ext-0.187
 how-0.173
Token sleepy
Token ID42023
Feature activation-2.47e-01

Pos logit contributions
 faded+0.300
mail+0.284
 shone+0.273
 tangled+0.272
 melted+0.270
Neg logit contributions
 Amen-0.301
ric-0.280
iu-0.276
 Wed-0.271
 how-0.269
Token.
Token ID13
Feature activation-1.18e-01

Pos logit contributions
 tangled+0.365
 faded+0.365
 shone+0.348
 melted+0.348
mail+0.346
Neg logit contributions
 Amen-0.167
iu-0.160
 how-0.137
ric-0.136
ext-0.133
Token He
Token ID679
Feature activation-2.09e-01

Pos logit contributions
 faded+0.145
 melted+0.137
 tangled+0.135
 shone+0.133
 disappeared+0.130
Neg logit contributions
ric-0.116
iu-0.115
 Amen-0.115
ext-0.104
 interests-0.102
Token yawn
Token ID40908
Feature activation-2.49e-01

Pos logit contributions
 tangled+0.244
Ancient+0.240
Factory+0.235
mail+0.233
 triggers+0.233
Neg logit contributions
 Amen-0.193
 spend-0.190
 how-0.185
-0.181
 go-0.178
Tokened
Token ID276
Feature activation-3.51e-01

Pos logit contributions
 swarm+0.290
 tangled+0.280
Bee+0.271
 swollen+0.270
 rotten+0.269
Neg logit contributions
 how-0.251
 as-0.249
 that-0.227
 about-0.222
 Just-0.216
Token,
Token ID11
Feature activation-1.75e-01

Pos logit contributions
Factory+0.578
 leaked+0.574
mail+0.573
 tangled+0.571
 faded+0.559
Neg logit contributions
 how-0.218
 Amen-0.172
 why-0.168
 where-0.163
 as-0.157
Token rubbing
Token ID27714
Feature activation-2.93e-01

Pos logit contributions
Factory+0.221
mail+0.219
eless+0.216
 iv+0.216
Ancient+0.215
Neg logit contributions
 how-0.162
 where-0.138
 why-0.137
 Amen-0.136
iu-0.134
Token his
Token ID465
Feature activation-2.88e-01

Pos logit contributions
 faded+0.356
 tangled+0.332
 melted+0.328
 shone+0.319
 flowed+0.317
Neg logit contributions
 Amen-0.293
iu-0.292
ric-0.287
ext-0.271
 interests-0.259
Token eyes
Token ID2951
Feature activation-2.90e-01

Pos logit contributions
 faded+0.324
 tangled+0.303
 shone+0.298
 melted+0.296
 flowed+0.290
Neg logit contributions
iu-0.314
 Amen-0.312
ric-0.305
ext-0.295
 interests-0.279
Token.
Token ID13
Feature activation-1.31e-01

Pos logit contributions
Factory+0.500
mail+0.489
Ancient+0.487
Dead+0.484
Spider+0.483
Neg logit contributions
 as-0.131
 how-0.115
 where-0.086
 go-0.085
 that-0.075
Token.
Token ID13
Feature activation-1.16e-01

Pos logit contributions
 tangled+0.388
 flowed+0.385
 faded+0.383
Bee+0.381
 stiff+0.372
Neg logit contributions
iu-0.146
 Amen-0.138
 how-0.136
 interests-0.130
pieces-0.112
Token
Token ID198
Feature activation-1.00e-01

Pos logit contributions
 faded+0.133
 tangled+0.127
 shone+0.125
 melted+0.123
 flowed+0.122
Neg logit contributions
iu-0.124
 Amen-0.122
ric-0.118
ext-0.113
 interests-0.104
Token
Token ID198
Feature activation-1.02e-01

Pos logit contributions
 faded+0.135
 tangled+0.127
 melted+0.127
 shone+0.127
 healed+0.123
Neg logit contributions
iu-0.088
 Amen-0.087
ric-0.086
ext-0.080
 interests-0.075
TokenThey
Token ID2990
Feature activation-1.84e-01

Pos logit contributions
 faded+0.122
 tangled+0.114
 melted+0.114
 shone+0.113
 healed+0.109
Neg logit contributions
iu-0.105
 Amen-0.104
ric-0.103
ext-0.096
 interests-0.091
Token sne
Token ID10505
Feature activation-2.67e-01

Pos logit contributions
 tangled+0.213
 faded+0.209
 darker+0.202
iest+0.202
 flowed+0.202
Neg logit contributions
 spend-0.178
 Amen-0.174
 wait-0.171
iu-0.166
 how-0.159
Tokenaked
Token ID4335
Feature activation-3.50e-01

Pos logit contributions
 faded+0.208
 tangled+0.189
 shone+0.187
 flowed+0.182
 dim+0.181
Neg logit contributions
iu-0.384
 Amen-0.378
 interests-0.376
ric-0.371
ext-0.361
Token into
Token ID656
Feature activation-2.53e-01

Pos logit contributions
 faded+0.528
 shone+0.521
 flowed+0.504
 healed+0.500
chy+0.499
Neg logit contributions
iu-0.249
 how-0.221
 Amen-0.219
ext-0.213
 Wed-0.189
Token their
Token ID511
Feature activation-2.50e-01

Pos logit contributions
 faded+0.329
 shone+0.315
chy+0.314
 healed+0.306
 melted+0.305
Neg logit contributions
iu-0.230
 Amen-0.225
ric-0.218
ext-0.202
 how-0.201
Token room
Token ID2119
Feature activation-2.74e-01

Pos logit contributions
 faded+0.283
 shone+0.265
 tangled+0.264
 melted+0.262
 healed+0.256
Neg logit contributions
iu-0.270
 Amen-0.265
ric-0.261
ext-0.251
 interests-0.239
Token and
Token ID290
Feature activation-1.97e-01

Pos logit contributions
 tangled+0.441
 shone+0.436
 faded+0.431
Spider+0.427
 flowed+0.425
Neg logit contributions
 where-0.142
 how-0.140
iu-0.137
 interests-0.118
 Amen-0.116
Token took
Token ID1718
Feature activation-2.96e-01

Pos logit contributions
 faded+0.183
 tangled+0.166
Factory+0.166
 iv+0.165
 healed+0.165
Neg logit contributions
iu-0.238
 Amen-0.232
 how-0.221
ric-0.214
ext-0.212
Token someone
Token ID2130
Feature activation-2.68e-01

Pos logit contributions
 faded+0.360
 tangled+0.342
chy+0.335
 shone+0.332
Factory+0.329
Neg logit contributions
 how-0.200
iu-0.190
 Amen-0.179
ext-0.178
 what-0.174
Token who
Token ID508
Feature activation-1.91e-01

Pos logit contributions
 faded+0.319
 tangled+0.311
 melted+0.306
 shone+0.303
 healed+0.298
Neg logit contributions
iu-0.290
 Amen-0.290
ric-0.284
ext-0.256
icol-0.247
Token would
Token ID561
Feature activation-1.94e-01

Pos logit contributions
 swelling+0.231
 stiff+0.229
 faded+0.228
Factory+0.226
sheet+0.224
Neg logit contributions
 how-0.186
 you-0.167
 why-0.162
 what-0.159
 we-0.157
Token match
Token ID2872
Feature activation-3.02e-01

Pos logit contributions
 faded+0.290
 tangled+0.278
 shone+0.278
 melted+0.275
 flowed+0.268
Neg logit contributions
 Amen-0.136
iu-0.134
ric-0.132
ext-0.120
 interests-0.113
Token him
Token ID683
Feature activation-2.59e-01

Pos logit contributions
 faded+0.445
 melted+0.416
 shone+0.412
 tangled+0.411
 healed+0.406
Neg logit contributions
 how-0.226
ric-0.225
iu-0.202
 interests-0.193
ext-0.190
Token perfectly
Token ID7138
Feature activation-3.49e-01

Pos logit contributions
 faded+0.432
Spider+0.414
 flowed+0.408
 shone+0.408
Ancient+0.405
Neg logit contributions
 how-0.167
 interests-0.135
 where-0.122
ric-0.119
ap-0.110
Token.
Token ID13
Feature activation-1.18e-01

Pos logit contributions
 faded+0.541
Spider+0.532
 shone+0.527
mail+0.526
 flowed+0.523
Neg logit contributions
 how-0.191
iu-0.180
ric-0.174
 interests-0.171
ext-0.160
Token After
Token ID2293
Feature activation-2.06e-01

Pos logit contributions
 faded+0.182
 tangled+0.179
 shone+0.175
 melted+0.172
 dim+0.167
Neg logit contributions
iu-0.079
 Amen-0.078
ext-0.075
ric-0.075
 We-0.072
Token a
Token ID257
Feature activation-2.60e-01

Pos logit contributions
 tangled+0.303
chy+0.302
 faded+0.296
 shone+0.289
 stiff+0.284
Neg logit contributions
 how-0.154
iu-0.141
 what-0.135
 Amen-0.131
ric-0.128
Token while
Token ID981
Feature activation-2.34e-01

Pos logit contributions
 faded+0.265
 tangled+0.252
 shone+0.247
 melted+0.246
 healed+0.237
Neg logit contributions
iu-0.307
 Amen-0.303
ric-0.296
ext-0.284
 interests-0.273
Token,
Token ID11
Feature activation-1.69e-01

Pos logit contributions
chy+0.435
 shone+0.432
mail+0.430
 tangled+0.422
 iv+0.421
Neg logit contributions
iu-0.055
 as-0.051
 Amen-0.049
ap-0.048
 we-0.045
TokenThe
Token ID464
Feature activation-2.30e-01

Pos logit contributions
 faded+0.118
 tangled+0.112
 melted+0.110
 shone+0.110
 healed+0.106
Neg logit contributions
iu-0.112
 Amen-0.110
ric-0.109
ext-0.102
 interests-0.097
Token clown
Token ID25573
Feature activation-2.71e-01

Pos logit contributions
 faded+0.242
 tangled+0.231
 shone+0.223
chy+0.220
 healed+0.219
Neg logit contributions
 Amen-0.274
iu-0.270
ric-0.253
ext-0.252
 interests-0.238
Token smiled
Token ID13541
Feature activation-2.65e-01

Pos logit contributions
 tangled+0.402
 infected+0.401
 darker+0.400
Lost+0.399
 iv+0.399
Neg logit contributions
 how-0.228
 who-0.209
 then-0.185
 that-0.164
 each-0.162
Token and
Token ID290
Feature activation-1.92e-01

Pos logit contributions
 faded+0.308
 shone+0.283
 tangled+0.282
 melted+0.282
 healed+0.272
Neg logit contributions
iu-0.292
ric-0.288
 Amen-0.285
ext-0.264
 interests-0.252
Token wink
Token ID41955
Feature activation-2.87e-01

Pos logit contributions
Factory+0.203
chy+0.202
ooky+0.187
 iv+0.186
mail+0.185
Neg logit contributions
 how-0.227
iu-0.200
icol-0.188
ext-0.181
 where-0.177
Tokened
Token ID276
Feature activation-3.49e-01

Pos logit contributions
 swarm+0.349
Factory+0.328
Bee+0.328
lab+0.317
Spider+0.316
Neg logit contributions
 how-0.310
 as-0.261
 that-0.248
iu-0.247
 each-0.239
Token.
Token ID13
Feature activation-1.31e-01

Pos logit contributions
 faded+0.492
 tangled+0.479
 melted+0.473
 flowed+0.468
 shone+0.467
Neg logit contributions
iu-0.259
 Amen-0.253
ext-0.229
 how-0.228
ric-0.218
Token He
Token ID679
Feature activation-2.12e-01

Pos logit contributions
 faded+0.168
 tangled+0.163
 melted+0.163
 shone+0.160
chy+0.157
Neg logit contributions
 Amen-0.117
iu-0.115
ric-0.112
ext-0.108
 interests-0.103
Token took
Token ID1718
Feature activation-3.01e-01

Pos logit contributions
Factory+0.268
 darker+0.266
 weaker+0.256
 iv+0.252
 swarm+0.252
Neg logit contributions
 how-0.212
 who-0.174
 then-0.160
 each-0.159
ext-0.159
Token out
Token ID503
Feature activation-2.87e-01

Pos logit contributions
 faded+0.435
 shone+0.417
chy+0.415
 flowed+0.407
 tangled+0.402
Neg logit contributions
iu-0.217
 how-0.205
ext-0.200
 our-0.184
 Amen-0.179
Token another
Token ID1194
Feature activation-3.05e-01

Pos logit contributions
 faded+0.396
chy+0.390
 shone+0.374
 flowed+0.373
 melted+0.371
Neg logit contributions
 how-0.232
iu-0.216
ext-0.203
 Amen-0.200
ric-0.200
Token his
Token ID465
Feature activation-2.58e-01

Pos logit contributions
 faded+0.425
 shone+0.420
 flowed+0.413
 bothered+0.408
 spoke+0.407
Neg logit contributions
 how-0.125
 Wed-0.104
 where-0.097
iu-0.091
 each-0.089
Token car
Token ID1097
Feature activation-2.78e-01

Pos logit contributions
 faded+0.275
 tangled+0.255
 shone+0.254
 melted+0.254
 healed+0.243
Neg logit contributions
iu-0.300
 Amen-0.294
ric-0.292
ext-0.277
 interests-0.264
Token.
Token ID13
Feature activation-1.09e-01

Pos logit contributions
 faded+0.464
 swarm+0.455
 melted+0.444
 shone+0.442
Bee+0.436
Neg logit contributions
 how-0.149
 where-0.142
iu-0.133
 Amen-0.109
ext-0.109
Token It
Token ID632
Feature activation-2.27e-01

Pos logit contributions
 shone+0.164
 faded+0.164
 melted+0.162
 tangled+0.158
 wil+0.157
Neg logit contributions
iu-0.065
 how-0.063
 Amen-0.060
 We-0.059
ext-0.059
Token fit
Token ID4197
Feature activation-2.97e-01

Pos logit contributions
Lost+0.279
 swarm+0.272
Factory+0.272
Ancient+0.266
 poisonous+0.265
Neg logit contributions
 how-0.196
 go-0.185
ext-0.159
 just-0.156
iu-0.155
Token perfectly
Token ID7138
Feature activation-3.49e-01

Pos logit contributions
 swarm+0.531
 faded+0.484
Dead+0.480
Factory+0.477
Bee+0.475
Neg logit contributions
 how-0.206
 what-0.167
 where-0.164
 you-0.156
 as-0.151
Token.
Token ID13
Feature activation-1.17e-01

Pos logit contributions
 faded+0.538
Spider+0.525
 swarm+0.525
mail+0.522
Bee+0.522
Neg logit contributions
iu-0.205
 how-0.198
 where-0.171
ric-0.158
ext-0.156
Token He
Token ID679
Feature activation-1.98e-01

Pos logit contributions
 faded+0.168
 shone+0.167
 melted+0.164
 tangled+0.161
 spoke+0.158
Neg logit contributions
iu-0.082
ext-0.076
 Amen-0.076
 how-0.073
 We-0.072
Token was
Token ID373
Feature activation-2.42e-01

Pos logit contributions
 swarm+0.231
Bee+0.214
 darker+0.213
Factory+0.212
 iv+0.212
Neg logit contributions
 how-0.202
iu-0.188
 interests-0.174
ext-0.170
ric-0.162
Token happy
Token ID3772
Feature activation-2.56e-01

Pos logit contributions
Factory+0.353
Bee+0.348
mail+0.344
 iv+0.342
 faded+0.333
Neg logit contributions
 how-0.226
 where-0.193
 what-0.186
 why-0.177
 about-0.176
Token.
Token ID13
Feature activation-1.25e-01

Pos logit contributions
 faded+0.344
 tangled+0.326
 melted+0.324
 shone+0.323
 healed+0.315
Neg logit contributions
iu-0.224
 Amen-0.218
ric-0.216
ext-0.202
 interests-0.191
Token to
Token ID284
Feature activation-2.08e-01

Pos logit contributions
 melted+0.362
 faded+0.361
opy+0.360
chy+0.357
 leaked+0.355
Neg logit contributions
 how-0.134
 where-0.134
 Amen-0.126
 to-0.124
 as-0.120
Token them
Token ID606
Feature activation-2.70e-01

Pos logit contributions
 faded+0.210
 tangled+0.194
 melted+0.194
 shone+0.190
 healed+0.186
Neg logit contributions
iu-0.254
ric-0.249
 Amen-0.247
ext-0.234
 interests-0.228
Token.
Token ID13
Feature activation-1.21e-01

Pos logit contributions
 faded+0.458
 shone+0.438
 melted+0.434
 flowed+0.433
 tangled+0.421
Neg logit contributions
 Amen-0.140
 how-0.136
ric-0.133
iu-0.130
 where-0.117
Token It
Token ID632
Feature activation-2.37e-01

Pos logit contributions
 melted+0.187
 faded+0.186
 shone+0.182
chy+0.182
 tangled+0.179
Neg logit contributions
 Amen-0.073
ext-0.069
iu-0.067
ric-0.063
 We-0.063
Token sniff
Token ID26300
Feature activation-2.96e-01

Pos logit contributions
Factory+0.259
mail+0.247
Ancient+0.240
eless+0.239
ayer+0.230
Neg logit contributions
 how-0.250
 why-0.237
 Amen-0.235
ext-0.231
 where-0.222
Tokened
Token ID276
Feature activation-3.48e-01

Pos logit contributions
 crooked+0.328
 shadowy+0.328
 burnt+0.321
 swarm+0.320
 swollen+0.319
Neg logit contributions
 how-0.277
 each-0.261
 you-0.256
 what-0.256
 us-0.254
Token them
Token ID606
Feature activation-2.78e-01

Pos logit contributions
 faded+0.525
 flowed+0.516
 melted+0.516
Factory+0.514
chy+0.506
Neg logit contributions
 how-0.211
iu-0.203
 Amen-0.199
 why-0.180
ext-0.180
Token.
Token ID13
Feature activation-1.28e-01

Pos logit contributions
Ah+0.568
 swarm+0.568
Supp+0.568
 shadowy+0.565
 leaked+0.560
Neg logit contributions
 how-0.068
 as-0.053
ap-0.038
 where-0.029
 each-0.024
Token It
Token ID632
Feature activation-2.45e-01

Pos logit contributions
 melted+0.212
 faded+0.208
chy+0.207
 shone+0.206
 tangled+0.203
Neg logit contributions
 Amen-0.063
ext-0.061
 We-0.060
iu-0.055
 interests-0.050
Token l
Token ID300
Feature activation-2.57e-01

Pos logit contributions
Factory+0.276
mail+0.261
Ancient+0.255
Lost+0.245
Supp+0.243
Neg logit contributions
 how-0.255
 why-0.236
ext-0.236
 Amen-0.231
iu-0.226
Tokenicked
Token ID9484
Feature activation-2.67e-01

Pos logit contributions
 faded+0.319
 melted+0.317
 tangled+0.311
 dead+0.305
 shone+0.303
Neg logit contributions
iu-0.224
ext-0.216
ric-0.210
 Amen-0.204
ap-0.194
Token.
Token ID13
Feature activation-1.19e-01

Pos logit contributions
 faded+0.306
 tangled+0.289
 melted+0.289
 shone+0.288
 healed+0.280
Neg logit contributions
iu-0.189
 Amen-0.186
ric-0.183
ext-0.169
 interests-0.158
Token And
Token ID843
Feature activation-1.69e-01

Pos logit contributions
 faded+0.131
 shone+0.119
 melted+0.119
 healed+0.118
 tangled+0.118
Neg logit contributions
iu-0.134
ric-0.129
 Amen-0.128
ext-0.120
 interests-0.115
Token he
Token ID339
Feature activation-2.26e-01

Pos logit contributions
 faded+0.208
 shone+0.194
 melted+0.193
 tangled+0.193
 healed+0.187
Neg logit contributions
iu-0.169
 Amen-0.166
ric-0.165
ext-0.154
 interests-0.144
Token never
Token ID1239
Feature activation-2.49e-01

Pos logit contributions
 tangled+0.288
Ancient+0.281
 darker+0.278
Factory+0.278
Lost+0.274
Neg logit contributions
 how-0.217
 Amen-0.210
ric-0.198
ext-0.194
 Wed-0.183
Token yawn
Token ID40908
Feature activation-2.46e-01

Pos logit contributions
mail+0.257
Ancient+0.246
Factory+0.245
 tangled+0.241
 flowed+0.234
Neg logit contributions
ric-0.263
 how-0.258
 spend-0.254
 interests-0.247
iu-0.244
Tokened
Token ID276
Feature activation-3.48e-01

Pos logit contributions
 tangled+0.308
Bee+0.298
 buzz+0.295
 leaked+0.294
 swollen+0.293
Neg logit contributions
 how-0.197
!'-0.196
 about-0.188
ap-0.188
 Just-0.176
Token while
Token ID981
Feature activation-2.40e-01

Pos logit contributions
Factory+0.515
mail+0.512
 flowed+0.509
 tangled+0.496
 faded+0.495
Neg logit contributions
 how-0.239
iu-0.213
 why-0.212
 Amen-0.210
ric-0.209
Token singing
Token ID13777
Feature activation-2.72e-01

Pos logit contributions
mail+0.396
 iv+0.380
Factory+0.379
 tangled+0.378
aked+0.373
Neg logit contributions
 you-0.142
 how-0.135
 we-0.127
 what-0.127
 Wed-0.122
Token his
Token ID465
Feature activation-2.86e-01

Pos logit contributions
 tangled+0.440
mail+0.431
Factory+0.428
 leaked+0.422
Ancient+0.419
Neg logit contributions
 how-0.191
 why-0.145
 Amen-0.145
 as-0.139
 what-0.136
Token green
Token ID4077
Feature activation-2.80e-01

Pos logit contributions
 blocked+0.391
Supp+0.391
 flowed+0.390
Factory+0.390
mail+0.389
Neg logit contributions
 interests-0.238
ext-0.225
 Amen-0.216
iu-0.209
 how-0.208
Token song
Token ID3496
Feature activation-3.02e-01

Pos logit contributions
 tangled+0.357
Ancient+0.350
 fooled+0.344
mail+0.344
Factory+0.340
Neg logit contributions
ric-0.249
 interests-0.242
 Amen-0.232
 Wed-0.226
ext-0.220
Token stay
Token ID2652
Feature activation-2.76e-01

Pos logit contributions
 faded+0.231
 melted+0.226
 healed+0.222
 tangled+0.219
 shone+0.219
Neg logit contributions
ric-0.197
iu-0.195
 Amen-0.194
ext-0.185
 interests-0.169
Token there
Token ID612
Feature activation-2.21e-01

Pos logit contributions
 faded+0.276
 tangled+0.271
 shone+0.265
 melted+0.260
 healed+0.255
Neg logit contributions
iu-0.338
ric-0.338
 Amen-0.318
ext-0.309
 interests-0.294
Token.
Token ID13
Feature activation-1.06e-01

Pos logit contributions
 flowed+0.447
 leaked+0.441
 smoked+0.440
 faded+0.440
Spider+0.436
Neg logit contributions
 where-0.063
 how-0.054
 as-0.050
 about-0.036
 for-0.035
Token She
Token ID1375
Feature activation-1.88e-01

Pos logit contributions
 faded+0.149
 melted+0.141
 tangled+0.141
 shone+0.140
chy+0.138
Neg logit contributions
 Amen-0.086
iu-0.085
ric-0.081
ext-0.077
 Wed-0.074
Token fit
Token ID4197
Feature activation-2.95e-01

Pos logit contributions
 darker+0.209
 triggers+0.203
 tangled+0.202
 faded+0.200
 stiff+0.199
Neg logit contributions
 how-0.201
iu-0.182
ext-0.171
 why-0.161
 spend-0.159
Token perfectly
Token ID7138
Feature activation-3.48e-01

Pos logit contributions
 faded+0.479
 swarm+0.447
 flowed+0.444
Factory+0.444
 leaked+0.443
Neg logit contributions
 how-0.209
 you-0.179
 what-0.177
iu-0.176
 where-0.165
Token into
Token ID656
Feature activation-2.42e-01

Pos logit contributions
 faded+0.541
mail+0.538
Spider+0.535
Lost+0.534
Ah+0.527
Neg logit contributions
 how-0.218
iu-0.195
 as-0.184
 where-0.183
ast-0.158
Token the
Token ID262
Feature activation-2.47e-01

Pos logit contributions
 faded+0.379
chy+0.374
 spoke+0.361
 flowed+0.361
 shone+0.355
Neg logit contributions
 how-0.184
 what-0.157
 where-0.151
iu-0.135
 Wed-0.133
Token chair
Token ID5118
Feature activation-2.87e-01

Pos logit contributions
 faded+0.304
 flowed+0.295
 spoke+0.292
mail+0.285
Factory+0.277
Neg logit contributions
 Amen-0.233
iu-0.211
 Wed-0.202
 interests-0.189
 where-0.187
Token.
Token ID13
Feature activation-1.18e-01

Pos logit contributions
 flowed+0.473
Factory+0.472
mail+0.472
Ancient+0.468
 faded+0.468
Neg logit contributions
 where-0.148
 how-0.133
 as-0.112
ap-0.107
 that-0.103
Token
Token ID198
Feature activation-1.05e-01

Pos logit contributions
 faded+0.175
 melted+0.167
 tangled+0.166
 shone+0.163
chy+0.162
Neg logit contributions
 Amen-0.089
iu-0.087
ric-0.085
ext-0.080
 Wed-0.075
Tokenheet
Token ID25473
Feature activation-2.68e-01

Pos logit contributions
 faded+0.340
 shone+0.330
 flowed+0.323
 tangled+0.322
mail+0.318
Neg logit contributions
iu-0.225
 Amen-0.214
ext-0.202
 how-0.202
 Wed-0.201
Token.
Token ID13
Feature activation-1.05e-01

Pos logit contributions
 faded+0.376
 shone+0.363
 tangled+0.363
 flowed+0.355
 melted+0.354
Neg logit contributions
iu-0.202
 Amen-0.192
ric-0.182
 Wed-0.175
ext-0.175
Token He
Token ID679
Feature activation-1.91e-01

Pos logit contributions
 faded+0.143
 melted+0.135
 tangled+0.135
 shone+0.134
chy+0.130
Neg logit contributions
 Amen-0.090
iu-0.090
ric-0.088
ext-0.081
 interests-0.076
Token did
Token ID750
Feature activation-2.31e-01

Pos logit contributions
 darker+0.244
 iv+0.241
 tangled+0.234
Lost+0.233
 triggers+0.232
Neg logit contributions
 how-0.179
-0.150
iu-0.149
ric-0.139
 interests-0.137
Token it
Token ID340
Feature activation-2.69e-01

Pos logit contributions
 faded+0.300
 tangled+0.283
 melted+0.282
 shone+0.281
 healed+0.273
Neg logit contributions
iu-0.212
 Amen-0.210
ric-0.206
ext-0.192
 interests-0.181
Token perfectly
Token ID7138
Feature activation-3.47e-01

Pos logit contributions
Bee+0.422
 iv+0.422
 poisoned+0.419
 tangled+0.418
 faded+0.418
Neg logit contributions
 how-0.162
 Wed-0.142
 spend-0.125
ext-0.121
!'-0.119
Token!
Token ID0
Feature activation-1.25e-01

Pos logit contributions
Bee+0.614
Spider+0.592
mail+0.586
Lost+0.580
Fox+0.576
Neg logit contributions
 how-0.179
 Wed-0.126
 as-0.120
!'-0.118
 where-0.108
Token
Token ID198
Feature activation-9.93e-02

Pos logit contributions
 faded+0.163
 tangled+0.154
 melted+0.153
 shone+0.153
 healed+0.148
Neg logit contributions
iu-0.114
 Amen-0.112
ric-0.110
ext-0.103
 interests-0.097
Token
Token ID198
Feature activation-1.01e-01

Pos logit contributions
 faded+0.139
 melted+0.131
 tangled+0.130
 shone+0.130
 flowed+0.126
Neg logit contributions
 Amen-0.083
iu-0.082
ric-0.080
ext-0.075
 interests-0.069
TokenThe
Token ID464
Feature activation-2.29e-01

Pos logit contributions
 faded+0.130
 melted+0.123
 shone+0.122
 tangled+0.121
chy+0.119
Neg logit contributions
 Amen-0.095
iu-0.094
ric-0.092
ext-0.087
 interests-0.081
Token tutor
Token ID48680
Feature activation-2.79e-01

Pos logit contributions
 faded+0.304
 tangled+0.291
 shone+0.289
 flowed+0.284
 melted+0.283
Neg logit contributions
 Amen-0.210
iu-0.204
ext-0.185
ric-0.181
 interests-0.176
Token him
Token ID683
Feature activation-2.51e-01

Pos logit contributions
 faded+0.337
 tangled+0.331
 shone+0.321
chy+0.316
 flowed+0.314
Neg logit contributions
 Amen-0.149
iu-0.143
ext-0.127
ric-0.124
 how-0.124
Token and
Token ID290
Feature activation-1.78e-01

Pos logit contributions
Spider+0.473
Bee+0.472
 faded+0.470
 swarm+0.469
 darker+0.463
Neg logit contributions
 how-0.120
 where-0.091
 interests-0.077
 what-0.073
 why-0.073
Token it
Token ID340
Feature activation-2.67e-01

Pos logit contributions
 faded+0.180
 melted+0.169
 shone+0.169
 tangled+0.165
 healed+0.161
Neg logit contributions
iu-0.220
ric-0.218
 Amen-0.212
ext-0.202
 interests-0.190
Token fit
Token ID4197
Feature activation-2.94e-01

Pos logit contributions
 swarm+0.488
Bee+0.486
Spider+0.456
Ancient+0.451
Lost+0.450
Neg logit contributions
 how-0.147
 go-0.120
-0.111
 we-0.095
 that-0.093
Token him
Token ID683
Feature activation-2.57e-01

Pos logit contributions
 faded+0.444
 swarm+0.432
Dead+0.422
Bee+0.414
 buzz+0.412
Neg logit contributions
 how-0.207
iu-0.188
ric-0.179
 interests-0.179
ext-0.173
Token perfectly
Token ID7138
Feature activation-3.47e-01

Pos logit contributions
 swarm+0.404
 faded+0.398
Bee+0.397
Spider+0.393
Dead+0.390
Neg logit contributions
 how-0.179
 where-0.147
 interests-0.146
ext-0.137
 what-0.135
Token.
Token ID13
Feature activation-1.09e-01

Pos logit contributions
 faded+0.527
Bee+0.523
 swarm+0.518
Spider+0.517
mail+0.505
Neg logit contributions
iu-0.206
 how-0.188
ext-0.181
 interests-0.168
 Amen-0.166
Token He
Token ID679
Feature activation-2.00e-01

Pos logit contributions
 faded+0.119
 melted+0.113
 shone+0.112
 tangled+0.111
 healed+0.108
Neg logit contributions
iu-0.122
ric-0.119
 Amen-0.112
ext-0.109
 interests-0.105
Token looked
Token ID3114
Feature activation-2.66e-01

Pos logit contributions
 swarm+0.268
 darker+0.265
 iv+0.256
 weaker+0.247
 triggers+0.246
Neg logit contributions
 how-0.179
 interests-0.156
 spend-0.155
ext-0.149
 Amen-0.146
Token in
Token ID287
Feature activation-2.44e-01

Pos logit contributions
 faded+0.303
 tangled+0.287
 melted+0.286
 shone+0.284
 healed+0.273
Neg logit contributions
iu-0.294
ric-0.290
 Amen-0.287
ext-0.267
 interests-0.253
Token the
Token ID262
Feature activation-2.52e-01

Pos logit contributions
 faded+0.216
 tangled+0.208
 melted+0.208
 shone+0.197
 dark+0.193
Neg logit contributions
ric-0.337
iu-0.333
 Amen-0.332
ext-0.312
 interests-0.288
Token compassionate
Token ID32533
Feature activation-2.57e-01

Pos logit contributions
 iv+0.182
 faded+0.182
mail+0.181
 shone+0.172
Factory+0.172
Neg logit contributions
 Amen-0.212
iu-0.210
ext-0.200
ric-0.198
 interests-0.197
Token and
Token ID290
Feature activation-1.74e-01

Pos logit contributions
 faded+0.361
 tangled+0.354
 melted+0.346
 shone+0.345
 flowed+0.344
Neg logit contributions
 Amen-0.196
iu-0.191
ric-0.174
 how-0.172
 interests-0.170
Token kept
Token ID4030
Feature activation-2.82e-01

Pos logit contributions
 faded+0.178
 tangled+0.165
 melted+0.165
 shone+0.164
 healed+0.157
Neg logit contributions
iu-0.209
 Amen-0.206
ric-0.204
ext-0.194
 interests-0.185
Token the
Token ID262
Feature activation-2.41e-01

Pos logit contributions
 faded+0.314
 tangled+0.295
 shone+0.293
 melted+0.292
 healed+0.281
Neg logit contributions
iu-0.313
 Amen-0.308
ric-0.304
ext-0.287
 interests-0.272
Token rhythm
Token ID18662
Feature activation-2.79e-01

Pos logit contributions
mail+0.327
 faded+0.318
 flowed+0.310
 weighed+0.305
 bitten+0.300
Neg logit contributions
 interests-0.182
 Amen-0.177
iu-0.167
ap-0.167
ext-0.165
Token perfectly
Token ID7138
Feature activation-3.47e-01

Pos logit contributions
mail+0.415
 faded+0.399
Factory+0.395
 tangled+0.394
 flowed+0.386
Neg logit contributions
 how-0.177
 Amen-0.175
iu-0.159
 interests-0.157
 Wed-0.156
Token.
Token ID13
Feature activation-1.16e-01

Pos logit contributions
mail+0.546
Spider+0.540
Factory+0.537
Bee+0.533
 faded+0.532
Neg logit contributions
 how-0.178
iu-0.171
 Amen-0.154
!'-0.154
 interests-0.149
Token She
Token ID1375
Feature activation-1.93e-01

Pos logit contributions
 faded+0.158
 tangled+0.151
 melted+0.151
 shone+0.150
 healed+0.146
Neg logit contributions
iu-0.098
 Amen-0.096
ric-0.096
ext-0.090
 Wed-0.084
Token was
Token ID373
Feature activation-2.43e-01

Pos logit contributions
Factory+0.213
mail+0.211
 darker+0.208
 iv+0.206
 tangled+0.204
Neg logit contributions
 how-0.215
 interests-0.189
ext-0.180
 Amen-0.180
iu-0.176
Token so
Token ID523
Feature activation-1.93e-01

Pos logit contributions
 faded+0.306
mail+0.288
 shone+0.286
 melted+0.285
 tangled+0.284
Neg logit contributions
 Amen-0.232
ric-0.224
iu-0.222
ext-0.207
 how-0.203
Token proud
Token ID6613
Feature activation-2.20e-01

Pos logit contributions
 iv+0.279
mail+0.258
Factory+0.255
 triggers+0.239
 faded+0.238
Neg logit contributions
 Amen-0.170
 how-0.157
ext-0.145
ap-0.145
iu-0.137
Token found
Token ID1043
Feature activation-2.64e-01

Pos logit contributions
 swarm+0.225
 faded+0.222
 iv+0.220
 sting+0.218
 swollen+0.217
Neg logit contributions
 how-0.153
iu-0.152
 spend-0.150
ext-0.146
 why-0.137
Token one
Token ID530
Feature activation-2.43e-01

Pos logit contributions
 faded+0.411
chy+0.372
 swarm+0.360
 spoke+0.354
 dim+0.353
Neg logit contributions
 how-0.203
 what-0.168
iu-0.161
 where-0.160
ric-0.158
Token that
Token ID326
Feature activation-1.92e-01

Pos logit contributions
 faded+0.298
 healed+0.274
 tangled+0.271
 melted+0.270
 flowed+0.269
Neg logit contributions
ric-0.231
iu-0.226
 Amen-0.216
 how-0.214
ext-0.211
Token fit
Token ID4197
Feature activation-2.93e-01

Pos logit contributions
 faded+0.257
 swarm+0.254
 sting+0.249
opy+0.248
 poisonous+0.246
Neg logit contributions
 how-0.150
 Wed-0.135
 where-0.135
 we-0.129
 why-0.125
Token him
Token ID683
Feature activation-2.56e-01

Pos logit contributions
 faded+0.401
 swarm+0.381
Dead+0.375
Bee+0.372
 healed+0.372
Neg logit contributions
 how-0.232
ric-0.228
iu-0.221
 interests-0.211
ext-0.197
Token perfectly
Token ID7138
Feature activation-3.46e-01

Pos logit contributions
 swarm+0.464
Bee+0.435
Spider+0.423
 faded+0.419
Lost+0.409
Neg logit contributions
 how-0.201
 where-0.174
 what-0.153
 why-0.140
 as-0.138
Token and
Token ID290
Feature activation-1.82e-01

Pos logit contributions
Bee+0.626
 swarm+0.620
Spider+0.620
Mouse+0.605
Fox+0.599
Neg logit contributions
 how-0.161
 where-0.138
 as-0.125
iu-0.107
 what-0.105
Token it
Token ID340
Feature activation-2.73e-01

Pos logit contributions
 faded+0.208
Factory+0.200
 tangled+0.196
Bee+0.196
chy+0.196
Neg logit contributions
 how-0.190
iu-0.172
 interests-0.161
 why-0.160
 where-0.155
Token was
Token ID373
Feature activation-2.40e-01

Pos logit contributions
Lost+0.414
 swarm+0.411
Bee+0.407
Ancient+0.398
Factory+0.398
Neg logit contributions
 how-0.166
 just-0.151
-0.144
 go-0.139
n-0.138
Token dry
Token ID5894
Feature activation-2.84e-01

Pos logit contributions
pillar+0.423
Bee+0.402
Factory+0.394
Mouse+0.391
 dies+0.390
Neg logit contributions
 how-0.198
 where-0.162
 about-0.157
 as-0.149
 what-0.149
Token,
Token ID11
Feature activation-1.63e-01

Pos logit contributions
 faded+0.397
 shone+0.371
 tangled+0.369
 melted+0.367
 flowed+0.364
Neg logit contributions
iu-0.220
 Amen-0.214
ric-0.205
ext-0.195
 how-0.184
Token and
Token ID290
Feature activation-1.89e-01

Pos logit contributions
 flowed+0.462
Bee+0.452
 shone+0.445
 darker+0.445
 faded+0.440
Neg logit contributions
 how-0.168
 where-0.125
 what-0.122
 why-0.121
 about-0.116
Token forgot
Token ID16453
Feature activation-2.55e-01

Pos logit contributions
 faded+0.187
 shone+0.177
 melted+0.173
 tangled+0.170
 healed+0.168
Neg logit contributions
iu-0.243
ric-0.239
 Amen-0.231
ext-0.224
 interests-0.209
Token all
Token ID477
Feature activation-2.51e-01

Pos logit contributions
 faded+0.381
 tangled+0.363
 melted+0.361
 shone+0.360
 healed+0.351
Neg logit contributions
iu-0.185
 Amen-0.181
ric-0.178
ext-0.163
 interests-0.150
Token about
Token ID546
Feature activation-2.07e-01

Pos logit contributions
 faded+0.418
 tangled+0.399
 melted+0.398
 shone+0.397
 healed+0.388
Neg logit contributions
iu-0.142
 Amen-0.138
ric-0.135
ext-0.121
 interests-0.108
Token her
Token ID607
Feature activation-2.76e-01

Pos logit contributions
 faded+0.260
 shone+0.245
 tangled+0.243
 melted+0.242
 flowed+0.240
Neg logit contributions
iu-0.194
ric-0.192
 Amen-0.187
 Wed-0.179
ext-0.176
Token jealousy
Token ID35394
Feature activation-3.46e-01

Pos logit contributions
 faded+0.270
 shone+0.255
 flowed+0.254
 tangled+0.246
mail+0.242
Neg logit contributions
iu-0.331
 Amen-0.329
 interests-0.313
ext-0.313
ric-0.311
Token.
Token ID13
Feature activation-1.28e-01

Pos logit contributions
Bee+0.691
Spider+0.678
Ah+0.669
Mouse+0.650
Ancient+0.646
Neg logit contributions
 how-0.146
 about-0.077
 where-0.071
 as-0.069
 that-0.069
Token
Token ID198
Feature activation-1.13e-01

Pos logit contributions
 faded+0.121
 tangled+0.114
 shone+0.114
chy+0.112
 flowed+0.110
Neg logit contributions
iu-0.165
 Amen-0.156
ric-0.154
ext-0.143
 interests-0.138
Token
Token ID198
Feature activation-1.15e-01

Pos logit contributions
 faded+0.116
 tangled+0.112
 shone+0.109
 melted+0.108
 whispered+0.107
Neg logit contributions
iu-0.135
ric-0.128
 Amen-0.126
icol-0.114
 interests-0.113
TokenAfter
Token ID3260
Feature activation-2.27e-01

Pos logit contributions
 faded+0.131
 tangled+0.125
 melted+0.122
 shone+0.122
chy+0.119
Neg logit contributions
iu-0.126
ric-0.124
 Amen-0.122
ext-0.111
 interests-0.107
Token dinner
Token ID8073
Feature activation-2.90e-01

Pos logit contributions
 faded+0.336
 tangled+0.335
chy+0.331
 melted+0.324
mail+0.324
Neg logit contributions
 how-0.153
 Amen-0.148
iu-0.146
 Wed-0.145
 what-0.133
Token on
Token ID319
Feature activation-2.58e-01

Pos logit contributions
 faded+0.337
 melted+0.312
 tangled+0.311
 shone+0.309
 healed+0.303
Neg logit contributions
iu-0.273
 Amen-0.270
ric-0.266
ext-0.250
 interests-0.230
Token a
Token ID257
Feature activation-2.56e-01

Pos logit contributions
 faded+0.421
chy+0.408
 flowed+0.396
 bothered+0.395
 shone+0.395
Neg logit contributions
 how-0.146
 where-0.126
iu-0.125
 what-0.118
 Wed-0.116
Token bench
Token ID7624
Feature activation-2.68e-01

Pos logit contributions
 faded+0.176
 tangled+0.159
 melted+0.158
 shone+0.156
 healed+0.146
Neg logit contributions
iu-0.393
 Amen-0.389
ric-0.386
ext-0.370
 interests-0.359
Token and
Token ID290
Feature activation-1.92e-01

Pos logit contributions
 faded+0.351
 melted+0.330
 tangled+0.329
 shone+0.327
 healed+0.322
Neg logit contributions
iu-0.237
 Amen-0.234
ric-0.233
ext-0.214
 interests-0.199
Token peek
Token ID27185
Feature activation-2.87e-01

Pos logit contributions
 faded+0.187
chy+0.178
 weaker+0.175
Factory+0.173
 iv+0.171
Neg logit contributions
 how-0.218
iu-0.208
 where-0.194
icol-0.186
 what-0.186
Tokened
Token ID276
Feature activation-3.45e-01

Pos logit contributions
 faded+0.233
 swollen+0.223
 sore+0.223
 tangled+0.218
 healed+0.217
Neg logit contributions
iu-0.385
 how-0.359
icol-0.336
ext-0.332
!'-0.327
Token inside
Token ID2641
Feature activation-2.74e-01

Pos logit contributions
 faded+0.448
 melted+0.418
 tangled+0.417
 healed+0.415
 shone+0.411
Neg logit contributions
iu-0.308
 Amen-0.293
ric-0.288
 how-0.271
ext-0.268
Token.
Token ID13
Feature activation-1.26e-01

Pos logit contributions
 faded+0.407
 tangled+0.386
 melted+0.386
chy+0.383
 shone+0.383
Neg logit contributions
 Amen-0.186
iu-0.186
ric-0.177
 how-0.165
ext-0.162
Token
Token ID198
Feature activation-1.13e-01

Pos logit contributions
 faded+0.154
 tangled+0.145
 melted+0.144
 shone+0.144
 healed+0.139
Neg logit contributions
iu-0.126
 Amen-0.124
ric-0.123
ext-0.115
 interests-0.109
Token
Token ID198
Feature activation-1.14e-01

Pos logit contributions
 faded+0.141
 tangled+0.134
 melted+0.132
 shone+0.132
 healed+0.129
Neg logit contributions
iu-0.109
 Amen-0.106
ric-0.106
ext-0.097
 interests-0.092
Token"
Token ID1
Feature activation-1.99e-01

Pos logit contributions
 faded+0.137
 tangled+0.129
 melted+0.128
 shone+0.127
 healed+0.123
Neg logit contributions
iu-0.117
 Amen-0.115
ric-0.115
ext-0.107
 interests-0.101
Tokenmy
Token ID1820
Feature activation+5.23e-01

Pos logit contributions
chy+0.344
 faded+0.336
 healed+0.335
 shone+0.323
 melted+0.323
Neg logit contributions
 how-0.194
iu-0.184
 what-0.180
ric-0.169
 interests-0.167
Token,
Token ID11
Feature activation+5.19e-01

Pos logit contributions
 Amen+0.575
ric+0.569
iu+0.558
 interests+0.514
ext+0.513
Neg logit contributions
 faded-0.572
 tangled-0.540
 flowed-0.519
mail-0.518
 iv-0.510
Token Daddy
Token ID26926
Feature activation+3.84e-01

Pos logit contributions
 Amen+0.471
iu+0.468
ric+0.457
ext+0.420
 interests+0.409
Neg logit contributions
 faded-0.668
 tangled-0.628
 melted-0.621
 shone-0.617
mail-0.607
Token,
Token ID11
Feature activation+4.61e-01

Pos logit contributions
 Amen+0.361
ric+0.355
 how+0.339
iu+0.316
 why+0.306
Neg logit contributions
 faded-0.479
 tangled-0.467
 prick-0.462
 iv-0.461
mail-0.450
Token help
Token ID1037
Feature activation+3.63e-01

Pos logit contributions
 Amen+0.380
ric+0.371
iu+0.368
 how+0.339
 interests+0.337
Neg logit contributions
 faded-0.619
 tangled-0.593
mail-0.579
 shone-0.576
 melted-0.573
Token!
Token ID0
Feature activation+4.50e-01

Pos logit contributions
iu+0.261
ric+0.250
 Amen+0.243
ext+0.229
 interests+0.225
Neg logit contributions
 faded-0.546
 shone-0.519
 tangled-0.514
 melted-0.503
 flowed-0.501
Token I
Token ID314
Feature activation+3.63e-01

Pos logit contributions
iu+0.380
 Amen+0.380
ric+0.370
ext+0.342
 interests+0.320
Neg logit contributions
 faded-0.616
 tangled-0.584
 melted-0.583
 shone-0.579
 healed-0.560
Token lost
Token ID2626
Feature activation+3.06e-01

Pos logit contributions
 Amen+0.304
ric+0.250
 interests+0.240
 spend+0.238
iu+0.225
Neg logit contributions
 faded-0.493
mail-0.486
 tangled-0.481
Ancient-0.478
Factory-0.471
Token my
Token ID616
Feature activation+2.92e-01

Pos logit contributions
iu+0.195
 Amen+0.191
ric+0.187
ext+0.171
 interests+0.154
Neg logit contributions
 faded-0.483
 shone-0.457
 tangled-0.455
 melted-0.454
 flowed-0.447
Token marble
Token ID30623
Feature activation+2.48e-01

Pos logit contributions
ric+0.331
 Amen+0.328
iu+0.326
 interests+0.310
ext+0.307
Neg logit contributions
 faded-0.301
 flowed-0.281
mail-0.273
 spoke-0.271
 shone-0.270
Token!"
Token ID2474
Feature activation+3.88e-01

Pos logit contributions
iu+0.167
 Amen+0.156
ext+0.152
ric+0.146
 interests+0.136
Neg logit contributions
 faded-0.372
 shone-0.352
 flowed-0.350
 tangled-0.347
 spoke-0.340
Token tree
Token ID5509
Feature activation+4.36e-01

Pos logit contributions
 iv+0.222
 faded+0.217
chy+0.214
 tangled+0.210
 darker+0.205
Neg logit contributions
iu-0.216
 how-0.212
ric-0.206
ext-0.203
 interests-0.196
Token out
Token ID503
Feature activation+3.95e-01

Pos logit contributions
ric+0.341
 Amen+0.324
iu+0.318
 how+0.310
ext+0.306
Neg logit contributions
 faded-0.587
 tangled-0.555
 stiff-0.553
Factory-0.543
Ancient-0.543
Token of
Token ID286
Feature activation+4.43e-01

Pos logit contributions
 how+0.253
iu+0.243
 Amen+0.242
ric+0.238
ext+0.228
Neg logit contributions
mail-0.560
Ancient-0.551
 tangled-0.546
chy-0.545
 faded-0.544
Token the
Token ID262
Feature activation+4.07e-01

Pos logit contributions
 Amen+0.414
iu+0.406
ric+0.403
 how+0.384
ext+0.380
Neg logit contributions
 faded-0.539
chy-0.513
 shone-0.499
 flowed-0.491
 tangled-0.489
Token ground
Token ID2323
Feature activation+3.61e-01

Pos logit contributions
ric+0.464
 Amen+0.445
iu+0.436
 interests+0.406
 Wed+0.405
Neg logit contributions
 faded-0.423
 flowed-0.393
mail-0.385
 shone-0.383
 tangled-0.378
Token.
Token ID13
Feature activation+4.69e-01

Pos logit contributions
ric+0.187
 how+0.175
 where+0.159
!'+0.150
iu+0.139
Neg logit contributions
mail-0.561
 faded-0.557
Ancient-0.553
Factory-0.553
Spider-0.551
Token Mike
Token ID4995
Feature activation+3.04e-01

Pos logit contributions
 Amen+0.366
ric+0.324
iu+0.322
 We+0.306
ext+0.300
Neg logit contributions
 faded-0.701
 melted-0.686
 tangled-0.670
 shone-0.662
chy-0.654
Token was
Token ID373
Feature activation+3.25e-01

Pos logit contributions
ric+0.292
+0.256
 how+0.250
ast+0.228
!'+0.225
Neg logit contributions
 lighter-0.399
 darker-0.398
 faded-0.392
Lost-0.375
 thinner-0.375
Token so
Token ID523
Feature activation+3.71e-01

Pos logit contributions
ric+0.224
 how+0.217
 Amen+0.215
iu+0.194
 where+0.182
Neg logit contributions
mail-0.486
 faded-0.476
Factory-0.464
 iv-0.458
 flowed-0.451
Token happy
Token ID3772
Feature activation+3.01e-01

Pos logit contributions
 Amen+0.335
 how+0.309
+0.290
ext+0.288
iu+0.285
Neg logit contributions
mail-0.494
 iv-0.490
Factory-0.464
 faded-0.455
 triggers-0.438
Token and
Token ID290
Feature activation+3.53e-01

Pos logit contributions
 Amen+0.314
iu+0.304
ric+0.304
ext+0.274
 interests+0.259
Neg logit contributions
 faded-0.376
 melted-0.352
 shone-0.350
 tangled-0.341
 disappeared-0.340
Token home
Token ID1363
Feature activation+3.85e-01

Pos logit contributions
ric+0.337
 how+0.333
 interests+0.307
 you+0.306
ext+0.304
Neg logit contributions
Ancient-0.569
Factory-0.559
 faded-0.558
 White-0.556
mail-0.555
Token to
Token ID284
Feature activation+4.31e-01

Pos logit contributions
iu+0.299
ric+0.298
 Amen+0.292
ext+0.274
 interests+0.270
Neg logit contributions
 faded-0.531
 tangled-0.504
 shone-0.498
 melted-0.493
 flowed-0.490
Token play
Token ID711
Feature activation+3.51e-01

Pos logit contributions
iu+0.392
 Amen+0.388
ric+0.383
ext+0.356
 interests+0.335
Neg logit contributions
 faded-0.561
 tangled-0.531
 shone-0.527
 melted-0.527
 healed-0.508
Token.
Token ID13
Feature activation+4.61e-01

Pos logit contributions
 Amen+0.232
ext+0.222
 how+0.213
iu+0.209
 interests+0.196
Neg logit contributions
 faded-0.537
 tangled-0.505
 melted-0.499
 shone-0.495
 stiff-0.492
Token
Token ID198
Feature activation+4.71e-01

Pos logit contributions
 Amen+0.405
iu+0.400
ric+0.392
ext+0.366
 interests+0.344
Neg logit contributions
 faded-0.619
 melted-0.588
 tangled-0.586
 shone-0.581
 healed-0.563
Token
Token ID198
Feature activation+4.63e-01

Pos logit contributions
 Amen+0.376
iu+0.364
ric+0.361
ext+0.347
 interests+0.316
Neg logit contributions
 faded-0.679
 melted-0.646
 shone-0.640
 tangled-0.638
 healed-0.620
TokenSo
Token ID2396
Feature activation+4.05e-01

Pos logit contributions
 Amen+0.436
iu+0.411
ext+0.404
ric+0.401
 interests+0.374
Neg logit contributions
 faded-0.616
 melted-0.592
 shone-0.583
 tangled-0.573
 healed-0.564
Token she
Token ID673
Feature activation+3.32e-01

Pos logit contributions
iu+0.329
 Amen+0.318
ric+0.300
ext+0.294
 how+0.278
Neg logit contributions
 faded-0.559
 tangled-0.542
 melted-0.529
 shone-0.524
chy-0.521
Token asked
Token ID1965
Feature activation+3.36e-01

Pos logit contributions
 how+0.337
iu+0.336
 Amen+0.320
ext+0.314
 spend+0.313
Neg logit contributions
 faded-0.341
 tangled-0.338
 darker-0.335
 rotten-0.325
 dead-0.324
Token her
Token ID607
Feature activation+2.67e-01

Pos logit contributions
 how+0.274
 why+0.213
 what+0.211
 where+0.188
 each+0.167
Neg logit contributions
 faded-0.573
 tangled-0.561
 melted-0.548
chy-0.541
 leaked-0.536
Token mom
Token ID1995
Feature activation+2.70e-01

Pos logit contributions
 how+0.167
 why+0.140
 interests+0.136
iu+0.131
 what+0.129
Neg logit contributions
 faded-0.444
 tangled-0.429
 melted-0.425
 flowed-0.424
 shone-0.417
Token want
Token ID765
Feature activation+4.08e-01

Pos logit contributions
iu+0.451
ric+0.449
ext+0.437
 Amen+0.437
icol+0.399
Neg logit contributions
 faded-0.530
 melted-0.506
chy-0.498
 tangled-0.495
 healed-0.490
Token to
Token ID284
Feature activation+4.02e-01

Pos logit contributions
iu+0.287
 Amen+0.279
ric+0.276
ext+0.253
 interests+0.234
Neg logit contributions
 faded-0.617
 tangled-0.586
 shone-0.585
 melted-0.584
 healed-0.568
Token leave
Token ID2666
Feature activation+2.96e-01

Pos logit contributions
iu+0.334
 Amen+0.326
ric+0.321
ext+0.301
 interests+0.274
Neg logit contributions
 faded-0.562
 melted-0.534
 tangled-0.529
 shone-0.522
 healed-0.517
Token.
Token ID13
Feature activation+4.42e-01

Pos logit contributions
iu+0.267
 Amen+0.265
ric+0.258
ext+0.247
 how+0.236
Neg logit contributions
 faded-0.384
 shone-0.358
 melted-0.352
 tangled-0.350
 flowed-0.350
Token
Token ID198
Feature activation+4.49e-01

Pos logit contributions
iu+0.556
ric+0.527
 Amen+0.524
ext+0.503
 interests+0.482
Neg logit contributions
 faded-0.418
 tangled-0.398
 shone-0.383
 melted-0.380
 whispered-0.377
Token
Token ID198
Feature activation+4.41e-01

Pos logit contributions
iu+0.595
ric+0.554
 Amen+0.544
icol+0.529
pieces+0.499
Neg logit contributions
 tangled-0.391
 faded-0.387
mail-0.374
 whispered-0.371
 dead-0.353
TokenThe
Token ID464
Feature activation+3.14e-01

Pos logit contributions
iu+0.600
ric+0.589
 Amen+0.545
icol+0.536
ext+0.509
Neg logit contributions
 tangled-0.371
 faded-0.361
mail-0.344
Bee-0.343
 whispered-0.341
Token rain
Token ID6290
Feature activation+2.73e-01

Pos logit contributions
 Amen+0.341
iu+0.319
 interests+0.289
ric+0.286
ext+0.279
Neg logit contributions
 faded-0.343
mail-0.330
 shone-0.322
 flowed-0.320
 healed-0.318
Token started
Token ID2067
Feature activation+2.84e-01

Pos logit contributions
iu+0.249
ext+0.238
 how+0.234
 interests+0.230
ric+0.222
Neg logit contributions
 Berry-0.311
mail-0.299
Ancient-0.299
Bee-0.298
Pr-0.290
Token to
Token ID284
Feature activation+3.10e-01

Pos logit contributions
 how+0.031
 to+0.028
 where+0.015
ric+0.015
 as+0.009
Neg logit contributions
mail-0.602
wy-0.599
Mouse-0.592
Spider-0.584
Bee-0.579
Token fall
Token ID2121
Feature activation+2.44e-01

Pos logit contributions
ric+0.247
 Amen+0.237
 go+0.222
 how+0.221
 spend+0.220
Neg logit contributions
ed-0.427
mail-0.413
aked-0.411
 faded-0.409
able-0.404
Token that
Token ID326
Feature activation+5.46e-01

Pos logit contributions
 dim+0.435
 darker+0.426
Factory+0.420
 faded+0.418
Ancient+0.418
Neg logit contributions
-0.191
 how-0.186
n-0.180
 why-0.169
 just-0.166
Token were
Token ID547
Feature activation+4.54e-01

Pos logit contributions
 how+0.455
 you+0.452
 Amen+0.440
ric+0.422
 we+0.413
Neg logit contributions
mail-0.678
Factory-0.678
 faded-0.674
Ancient-0.671
 tangled-0.646
Token all
Token ID477
Feature activation+4.02e-01

Pos logit contributions
 how+0.433
 what+0.365
 why+0.363
 where+0.344
 our+0.341
Neg logit contributions
Factory-0.627
mail-0.617
Ancient-0.607
Spider-0.582
lab-0.575
Token mixed
Token ID7668
Feature activation+3.41e-01

Pos logit contributions
 how+0.405
 where+0.340
 you+0.335
 about+0.334
 our+0.331
Neg logit contributions
Spider-0.510
mail-0.503
Factory-0.481
 Berry-0.480
Ancient-0.477
Token up
Token ID510
Feature activation+3.44e-01

Pos logit contributions
iu+0.358
 Amen+0.352
ric+0.349
ext+0.328
 interests+0.312
Neg logit contributions
 faded-0.399
 tangled-0.373
 melted-0.372
 shone-0.370
 healed-0.358
Token.
Token ID13
Feature activation+4.65e-01

Pos logit contributions
iu+0.303
ric+0.299
 how+0.295
 interests+0.279
 Wed+0.279
Neg logit contributions
 faded-0.445
mail-0.409
 shone-0.405
 healed-0.401
 tangled-0.399
Token Jane
Token ID12091
Feature activation+2.88e-01

Pos logit contributions
 Amen+0.344
ric+0.319
iu+0.315
ext+0.291
 Wed+0.286
Neg logit contributions
 faded-0.708
 melted-0.685
 tangled-0.676
 shone-0.666
chy-0.659
Token didn
Token ID1422
Feature activation+3.30e-01

Pos logit contributions
 how+0.259
ric+0.248
+0.247
!'+0.216
iu+0.211
Neg logit contributions
 darker-0.377
Lost-0.366
mail-0.361
 faded-0.357
 iv-0.349
Token't
Token ID470
Feature activation+3.45e-01

Pos logit contributions
iu+0.282
 Amen+0.278
ric+0.274
ext+0.255
 interests+0.237
Neg logit contributions
 faded-0.448
 tangled-0.423
 melted-0.422
 shone-0.421
 healed-0.408
Token know
Token ID760
Feature activation+3.00e-01

Pos logit contributions
iu+0.269
 Amen+0.263
ric+0.254
 interests+0.231
ext+0.220
Neg logit contributions
 faded-0.488
 shone-0.473
 melted-0.464
 tangled-0.462
mail-0.455
Token what
Token ID644
Feature activation+3.10e-01

Pos logit contributions
iu+0.104
ric+0.099
 Amen+0.091
ext+0.077
 interests+0.067
Neg logit contributions
 faded-0.558
 melted-0.534
 tangled-0.534
 shone-0.533
 healed-0.525
Token shouted
Token ID17293
Feature activation+3.63e-01

Pos logit contributions
 how+0.300
+0.282
 Amen+0.282
ric+0.277
 you+0.273
Neg logit contributions
Ancient-0.417
mail-0.414
Factory-0.401
 darker-0.393
Lost-0.392
Token with
Token ID351
Feature activation+3.38e-01

Pos logit contributions
 how+0.197
ext+0.171
ap+0.169
 why+0.165
 what+0.163
Neg logit contributions
mail-0.599
 tangled-0.593
 shone-0.593
 faded-0.584
Factory-0.578
Token happiness
Token ID12157
Feature activation+2.78e-01

Pos logit contributions
iu+0.445
ric+0.445
 Amen+0.432
ext+0.404
 interests+0.392
Neg logit contributions
 faded-0.317
 melted-0.292
 tangled-0.291
 shone-0.282
 healed-0.278
Token and
Token ID290
Feature activation+3.72e-01

Pos logit contributions
ric+0.177
 how+0.175
iu+0.173
 Amen+0.166
 interests+0.161
Neg logit contributions
 tangled-0.403
 faded-0.400
mail-0.393
Ancient-0.391
Factory-0.384
Token the
Token ID262
Feature activation+3.13e-01

Pos logit contributions
 how+0.404
iu+0.374
 Amen+0.363
 why+0.346
 what+0.345
Neg logit contributions
Factory-0.388
Ancient-0.381
chy-0.377
mail-0.375
 tangled-0.365
Token p
Token ID279
Feature activation+3.21e-01

Pos logit contributions
 Amen+0.391
iu+0.347
 interests+0.334
ric+0.326
 you+0.302
Neg logit contributions
 faded-0.307
 flowed-0.302
 tangled-0.283
mail-0.279
 shone-0.276
Tokenanda
Token ID5282
Feature activation+3.02e-01

Pos logit contributions
iu+0.327
 Amen+0.324
ric+0.317
ext+0.299
 interests+0.283
Neg logit contributions
 faded-0.385
 tangled-0.362
 melted-0.361
 shone-0.359
 healed-0.347
Token felt
Token ID2936
Feature activation+2.68e-01

Pos logit contributions
 how+0.285
 who+0.255
 as+0.243
iu+0.242
ric+0.239
Neg logit contributions
Ancient-0.370
 darker-0.363
Factory-0.356
Lost-0.353
 tangled-0.350
Token so
Token ID523
Feature activation+3.43e-01

Pos logit contributions
 Amen+0.266
 how+0.261
ext+0.242
 why+0.235
iu+0.232
Neg logit contributions
Factory-0.307
mail-0.307
 faded-0.291
Ancient-0.285
 shone-0.281
Token proud
Token ID6613
Feature activation+3.10e-01

Pos logit contributions
 Amen+0.394
ext+0.393
iu+0.376
 how+0.363
 interests+0.355
Neg logit contributions
mail-0.333
Factory-0.330
 iv-0.315
Ancient-0.304
 shone-0.303
Token.
Token ID13
Feature activation+3.90e-01

Pos logit contributions
iu+0.252
 Amen+0.244
ric+0.237
ext+0.224
 interests+0.212
Neg logit contributions
 faded-0.429
 tangled-0.412
 melted-0.408
 shone-0.406
 healed-0.398
TokenDo
Token ID5211
Feature activation+3.06e-01

Pos logit contributions
iu+0.385
 Amen+0.382
ric+0.377
ext+0.358
 interests+0.341
Neg logit contributions
 faded-0.330
 tangled-0.307
 melted-0.305
 shone-0.303
 healed-0.291
Token you
Token ID345
Feature activation+2.80e-01

Pos logit contributions
 Amen+0.416
icol+0.414
iu+0.391
ric+0.375
amb+0.352
Neg logit contributions
 weak-0.299
 short-0.298
 hurt-0.294
 small-0.288
 vanished-0.287
Token need
Token ID761
Feature activation+2.68e-01

Pos logit contributions
iu+0.254
 Amen+0.244
ric+0.243
ext+0.235
icol+0.209
Neg logit contributions
 faded-0.374
 melted-0.354
 tangled-0.351
 shone-0.350
 healed-0.346
Token help
Token ID1037
Feature activation+2.42e-01

Pos logit contributions
 Amen+0.150
 how+0.144
iu+0.137
ric+0.132
 Wed+0.125
Neg logit contributions
 faded-0.431
 tangled-0.414
chy-0.410
 shone-0.408
 flowed-0.408
Token?"
Token ID1701
Feature activation+3.31e-01

Pos logit contributions
iu+0.186
 Amen+0.182
ric+0.178
ext+0.165
 interests+0.152
Neg logit contributions
 faded-0.352
 tangled-0.336
 melted-0.335
 shone-0.333
 healed-0.326
Token Cotton
Token ID23915
Feature activation+2.19e-01

Pos logit contributions
 Amen+0.274
iu+0.232
ext+0.219
 Wed+0.207
ric+0.204
Neg logit contributions
 faded-0.475
 tangled-0.469
 melted-0.467
chy-0.457
 shone-0.457
Token said
Token ID531
Feature activation+2.63e-01

Pos logit contributions
 how+0.201
 Amen+0.176
iu+0.175
 why+0.173
 Just+0.162
Neg logit contributions
 tangled-0.298
 darker-0.287
Lost-0.277
 visible-0.274
mail-0.266
Token,
Token ID11
Feature activation+2.95e-01

Pos logit contributions
 how+0.127
 Amen+0.113
iu+0.104
 why+0.095
 where+0.092
Neg logit contributions
 tangled-0.460
 faded-0.454
mail-0.444
Factory-0.444
 melted-0.434
Token "
Token ID366
Feature activation+2.59e-01

Pos logit contributions
 how+0.166
 Amen+0.155
iu+0.128
 where+0.123
 why+0.117
Neg logit contributions
 tangled-0.529
able-0.509
 faded-0.507
mail-0.486
 leaked-0.484
TokenYes
Token ID5297
Feature activation+2.08e-01

Pos logit contributions
 Amen+0.183
iu+0.154
 interests+0.146
ext+0.134
ric+0.130
Neg logit contributions
 faded-0.404
 vanished-0.391
 melted-0.390
 spilled-0.388
 disappeared-0.387
Token,
Token ID11
Feature activation+2.75e-01

Pos logit contributions
iu+0.142
 Amen+0.138
ric+0.126
ext+0.122
 how+0.118
Neg logit contributions
 faded-0.317
 tangled-0.312
 shone-0.309
 flowed-0.302
mail-0.298
Token,
Token ID11
Feature activation+3.20e-01

Pos logit contributions
 how+0.163
ric+0.159
iu+0.146
ext+0.134
 Amen+0.130
Neg logit contributions
 faded-0.417
 shone-0.408
 tangled-0.400
 darker-0.395
 flowed-0.394
Token but
Token ID475
Feature activation+2.86e-01

Pos logit contributions
ric+0.328
 Amen+0.315
iu+0.315
 how+0.304
ext+0.290
Neg logit contributions
 faded-0.363
mail-0.357
Factory-0.344
 iv-0.339
 shone-0.339
Token G
Token ID402
Feature activation+2.44e-01

Pos logit contributions
iu+0.269
ric+0.263
 Amen+0.251
ext+0.240
 how+0.237
Neg logit contributions
 faded-0.351
 melted-0.336
 shone-0.329
 tangled-0.328
mail-0.325
Tokenogo
Token ID24076
Feature activation+2.39e-01

Pos logit contributions
iu+0.238
ric+0.234
 Amen+0.232
ext+0.218
 interests+0.204
Neg logit contributions
 faded-0.304
 tangled-0.286
 melted-0.284
 shone-0.283
 healed-0.274
Token dodged
Token ID49392
Feature activation+1.44e-01

Pos logit contributions
 how+0.175
n+0.160
 just+0.156
 why+0.141
+0.140
Neg logit contributions
Lost-0.352
 darker-0.335
Ancient-0.332
Factory-0.323
 lighter-0.322
Token.
Token ID13
Feature activation+3.29e-01

Pos logit contributions
 how+0.069
 what+0.063
iu+0.061
ric+0.058
 why+0.057
Neg logit contributions
 faded-0.237
 shone-0.230
Factory-0.229
Dead-0.228
 melted-0.227
Token He
Token ID679
Feature activation+2.43e-01

Pos logit contributions
ric+0.269
iu+0.249
icol+0.242
 Just+0.230
 Amen+0.221
Neg logit contributions
 melted-0.462
 faded-0.460
 tangled-0.447
chy-0.446
 shone-0.438
Token used
Token ID973
Feature activation+1.95e-01

Pos logit contributions
 how+0.221
 just+0.196
iu+0.195
ext+0.195
 why+0.189
Neg logit contributions
Factory-0.282
Ancient-0.282
mail-0.270
Lost-0.265
iest-0.264
Token his
Token ID465
Feature activation+1.63e-01

Pos logit contributions
 how+0.156
iu+0.128
ric+0.125
 interests+0.122
 what+0.116
Neg logit contributions
 faded-0.271
Factory-0.265
Ancient-0.265
chy-0.264
opy-0.263
Token horns
Token ID26771
Feature activation+1.71e-01

Pos logit contributions
 interests+0.154
ric+0.150
iu+0.146
ext+0.137
 how+0.133
Neg logit contributions
 faded-0.193
Ancient-0.193
 flowed-0.189
Factory-0.187
 shone-0.182
Token to
Token ID284
Feature activation+2.18e-01

Pos logit contributions
 how+0.063
ric+0.033
 interests+0.027
 where+0.027
 as+0.025
Neg logit contributions
 faded-0.312
Ancient-0.312
 swarm-0.311
mail-0.310
Bee-0.309
Token smiled
Token ID13541
Feature activation+2.32e-01

Pos logit contributions
 how+0.262
 why+0.237
iu+0.233
 Amen+0.233
ric+0.217
Neg logit contributions
mail-0.284
Lost-0.280
Factory-0.276
 tangled-0.273
Supp-0.258
Token,
Token ID11
Feature activation+3.24e-01

Pos logit contributions
iu+0.237
 Amen+0.233
ric+0.232
ext+0.217
 interests+0.205
Neg logit contributions
 faded-0.281
 tangled-0.262
 melted-0.261
 shone-0.261
 healed-0.252
Token "
Token ID366
Feature activation+2.84e-01

Pos logit contributions
 how+0.287
 Amen+0.280
iu+0.274
ric+0.264
 why+0.245
Neg logit contributions
mail-0.406
Factory-0.405
 faded-0.403
 tangled-0.401
chy-0.399
TokenYes
Token ID5297
Feature activation+2.36e-01

Pos logit contributions
 Amen+0.196
iu+0.178
ric+0.170
 interests+0.164
ext+0.151
Neg logit contributions
 faded-0.434
 melted-0.429
 spilled-0.419
chy-0.418
 vanished-0.416
Token I
Token ID314
Feature activation+2.57e-01

Pos logit contributions
iu+0.150
 Amen+0.135
ric+0.134
 how+0.130
ext+0.126
Neg logit contributions
 faded-0.369
 tangled-0.363
 shone-0.359
mail-0.354
 flowed-0.350
Token do
Token ID466
Feature activation+2.34e-01

Pos logit contributions
 Amen+0.214
iu+0.208
ric+0.204
 interests+0.189
ext+0.186
Neg logit contributions
 faded-0.351
 tangled-0.337
 melted-0.326
 shone-0.324
 stiff-0.317
Token.
Token ID13
Feature activation+3.28e-01

Pos logit contributions
iu+0.206
 Amen+0.205
ric+0.197
ext+0.184
 interests+0.175
Neg logit contributions
 faded-0.314
 tangled-0.297
 melted-0.295
 shone-0.294
 healed-0.284
Token Come
Token ID7911
Feature activation+2.18e-01

Pos logit contributions
iu+0.282
ric+0.266
ext+0.258
 Amen+0.248
 interests+0.222
Neg logit contributions
 faded-0.439
 shone-0.421
 tangled-0.407
 healed-0.404
 White-0.402
Token,
Token ID11
Feature activation+2.76e-01

Pos logit contributions
iu+0.263
ric+0.251
 Amen+0.248
ext+0.244
 interests+0.233
Neg logit contributions
 faded-0.219
 shone-0.211
 tangled-0.206
 melted-0.205
 healed-0.202
Token and
Token ID290
Feature activation+2.46e-01

Pos logit contributions
iu+0.208
 Amen+0.204
ric+0.198
ext+0.181
 interests+0.174
Neg logit contributions
 faded-0.398
 tangled-0.383
 shone-0.377
 melted-0.376
 flowed-0.370
Token I
Token ID314
Feature activation+2.20e-01

Pos logit contributions
iu+0.215
ric+0.208
ext+0.193
 Amen+0.184
 interests+0.170
Neg logit contributions
 faded-0.339
 shone-0.328
 whispered-0.325
 healed-0.319
 melted-0.319
Token didn
Token ID1422
Feature activation+3.50e-01

Pos logit contributions
 how+0.361
 spend+0.358
 Amen+0.357
 why+0.348
ric+0.333
Neg logit contributions
Factory-0.505
mail-0.500
 darker-0.491
Ancient-0.488
iest-0.484
Token't
Token ID470
Feature activation+3.68e-01

Pos logit contributions
iu+0.387
 Amen+0.368
ric+0.367
 interests+0.343
icol+0.324
Neg logit contributions
 faded-0.408
 tangled-0.394
 melted-0.379
chy-0.376
 healed-0.371
Token work
Token ID670
Feature activation+2.96e-01

Pos logit contributions
 Amen+0.247
iu+0.246
ric+0.232
 interests+0.222
 how+0.203
Neg logit contributions
 faded-0.556
 shone-0.544
 tangled-0.526
 melted-0.525
 flowed-0.522
Token anymore
Token ID7471
Feature activation+3.38e-01

Pos logit contributions
 how+0.188
 what+0.164
 about+0.152
 as+0.144
 where+0.144
Neg logit contributions
Bee-0.509
Spider-0.471
 shone-0.464
Mouse-0.464
mail-0.463
Token and
Token ID290
Feature activation+3.59e-01

Pos logit contributions
iu+0.240
ric+0.239
 Amen+0.235
 how+0.221
 interests+0.220
Neg logit contributions
 faded-0.493
 tangled-0.474
 shone-0.472
 melted-0.470
mail-0.463
Token she
Token ID673
Feature activation+3.24e-01

Pos logit contributions
 how+0.330
iu+0.311
 Amen+0.310
ric+0.308
 why+0.293
Neg logit contributions
mail-0.425
 faded-0.423
 tangled-0.418
Factory-0.417
 iv-0.409
Token felt
Token ID2936
Feature activation+2.57e-01

Pos logit contributions
 how+0.267
 Amen+0.256
ext+0.237
 spend+0.237
iu+0.237
Neg logit contributions
Factory-0.404
 darker-0.402
mail-0.399
Ancient-0.393
 iv-0.386
Token sad
Token ID6507
Feature activation+2.66e-01

Pos logit contributions
 Amen+0.301
 how+0.300
 why+0.280
 what+0.274
ric+0.271
Neg logit contributions
mail-0.255
 faded-0.240
Factory-0.240
 shone-0.238
Bee-0.231
Token.
Token ID13
Feature activation+3.88e-01

Pos logit contributions
 Amen+0.191
iu+0.190
ric+0.182
ext+0.176
 how+0.176
Neg logit contributions
 faded-0.375
 tangled-0.365
 shone-0.365
mail-0.365
 melted-0.361
Token
Token ID198
Feature activation+3.97e-01

Pos logit contributions
 Amen+0.310
ric+0.298
iu+0.298
ext+0.280
 interests+0.257
Neg logit contributions
 faded-0.558
 melted-0.541
 shone-0.533
 tangled-0.531
 healed-0.515
Token
Token ID198
Feature activation+3.92e-01

Pos logit contributions
 Amen+0.326
iu+0.320
ric+0.314
ext+0.302
 interests+0.276
Neg logit contributions
 faded-0.561
 melted-0.533
 shone-0.530
 tangled-0.526
 healed-0.511
Token".
Token ID1911
Feature activation+3.02e-01

Pos logit contributions
ric+0.091
 how+0.091
 Amen+0.085
iu+0.076
 you+0.072
Neg logit contributions
 faded-0.322
 shadowy-0.310
 tangled-0.309
Spider-0.309
 shone-0.308
Token
Token ID198
Feature activation+3.36e-01

Pos logit contributions
 Amen+0.243
iu+0.239
ric+0.232
ext+0.218
 interests+0.200
Neg logit contributions
 faded-0.434
 melted-0.414
 shone-0.410
 tangled-0.410
chy-0.398
Token
Token ID198
Feature activation+3.32e-01

Pos logit contributions
 Amen+0.282
iu+0.281
ric+0.275
ext+0.259
 interests+0.239
Neg logit contributions
 faded-0.466
 melted-0.440
 tangled-0.439
 shone-0.438
 flowed-0.423
TokenThe
Token ID464
Feature activation+2.05e-01

Pos logit contributions
 Amen+0.336
iu+0.321
ext+0.309
ric+0.309
 interests+0.291
Neg logit contributions
 faded-0.419
 melted-0.400
 shone-0.393
 tangled-0.387
chy-0.382
Token man
Token ID582
Feature activation+1.73e-01

Pos logit contributions
 Amen+0.217
iu+0.215
ric+0.197
 interests+0.195
 how+0.178
Neg logit contributions
mail-0.219
 faded-0.217
 tangled-0.215
 flowed-0.209
 pierced-0.209
Token agreed
Token ID4987
Feature activation+1.59e-01

Pos logit contributions
 how+0.166
 who+0.145
 why+0.134
 then+0.131
+0.128
Neg logit contributions
Lost-0.229
 darker-0.229
 tangled-0.228
chy-0.228
 prick-0.225
Token.
Token ID13
Feature activation+2.95e-01

Pos logit contributions
iu+0.134
 Amen+0.128
ric+0.120
ext+0.117
 how+0.113
Neg logit contributions
 faded-0.217
 tangled-0.210
 shone-0.208
 melted-0.204
 flowed-0.202
Token Together
Token ID17083
Feature activation+2.25e-01

Pos logit contributions
 Amen+0.246
iu+0.246
ric+0.234
 how+0.224
 We+0.214
Neg logit contributions
 faded-0.401
 melted-0.391
 shone-0.386
 tangled-0.386
chy-0.385
Token,
Token ID11
Feature activation+2.40e-01

Pos logit contributions
iu+0.125
 how+0.109
 Amen+0.105
ext+0.103
 Wed+0.093
Neg logit contributions
 tangled-0.358
 faded-0.356
 shone-0.352
 healed-0.349
mail-0.348
Token the
Token ID262
Feature activation+1.52e-01

Pos logit contributions
 how+0.101
iu+0.095
 we+0.079
ric+0.069
 We+0.066
Neg logit contributions
chy-0.417
 faded-0.416
 iv-0.406
mail-0.403
 shone-0.398
Token duck
Token ID22045
Feature activation+1.29e-01

Pos logit contributions
iu+0.163
 Amen+0.153
 interests+0.153
ric+0.143
 how+0.134
Neg logit contributions
 tangled-0.162
 faded-0.162
chy-0.158
mail-0.157
 pierced-0.153
Token owl
Token ID39610
Feature activation+8.24e-02

Pos logit contributions
 Amen+0.127
iu+0.118
 interests+0.114
ric+0.107
 how+0.102
Neg logit contributions
 faded-0.111
 flowed-0.108
 tangled-0.105
 sparked-0.103
 leaked-0.103
Token thought
Token ID1807
Feature activation+7.31e-02

Pos logit contributions
 how+0.084
 interests+0.073
n+0.070
 Amen+0.069
 then+0.068
Neg logit contributions
 darker-0.098
Factory-0.096
Ancient-0.095
 iv-0.095
eless-0.093
Token this
Token ID428
Feature activation+6.46e-02

Pos logit contributions
 how+0.045
 why+0.035
 what+0.035
 about+0.033
 where+0.031
Neg logit contributions
Factory-0.120
mail-0.120
Ancient-0.119
 shone-0.117
 faded-0.116
Token was
Token ID373
Feature activation+7.99e-02

Pos logit contributions
 interests+0.058
 how+0.057
 Amen+0.056
iu+0.051
 why+0.051
Neg logit contributions
 faded-0.078
 dim-0.075
Factory-0.074
Ancient-0.073
 tangled-0.072
Token a
Token ID257
Feature activation+6.15e-02

Pos logit contributions
 how+0.074
 why+0.064
 what+0.061
 where+0.056
 our+0.055
Neg logit contributions
Factory-0.111
 faded-0.109
mail-0.108
Ancient-0.105
Supp-0.103
Token good
Token ID922
Feature activation+4.60e-02

Pos logit contributions
 Amen+0.075
iu+0.070
ric+0.068
 interests+0.066
ext+0.066
Neg logit contributions
 faded-0.060
 tangled-0.058
 flowed-0.057
 shone-0.056
mail-0.055
Token idea
Token ID2126
Feature activation+6.20e-02

Pos logit contributions
ric+0.061
iu+0.060
 Amen+0.059
ext+0.056
 Wed+0.051
Neg logit contributions
 faded-0.044
 melted-0.041
 shone-0.040
 tangled-0.040
chy-0.038
Token and
Token ID290
Feature activation+1.20e-01

Pos logit contributions
iu+0.056
 Amen+0.054
ric+0.053
ext+0.050
 interests+0.048
Neg logit contributions
 faded-0.081
 tangled-0.077
 melted-0.076
 shone-0.076
 flowed-0.074
Token the
Token ID262
Feature activation+5.87e-02

Pos logit contributions
 how+0.120
iu+0.117
 Amen+0.116
ext+0.107
ric+0.103
Neg logit contributions
 faded-0.129
 iv-0.128
 tangled-0.128
chy-0.127
Factory-0.125
Token two
Token ID734
Feature activation+1.03e-02

Pos logit contributions
 Amen+0.066
 interests+0.060
iu+0.059
ric+0.053
 you+0.053
Neg logit contributions
 faded-0.062
 flowed-0.060
 tangled-0.057
 sparked-0.056
 leaked-0.056
Token friends
Token ID2460
Feature activation+5.87e-02

Pos logit contributions
 how+0.009
pieces+0.008
 interests+0.008
iu+0.008
 spend+0.008
Neg logit contributions
 iv-0.012
chy-0.012
Dead-0.012
 tangled-0.011
mail-0.011
Token it
Token ID340
Feature activation-1.28e-01

Pos logit contributions
 faded+0.083
mail+0.080
 tangled+0.079
 stiff+0.078
opy+0.078
Neg logit contributions
 how-0.040
 why-0.034
 what-0.030
ric-0.029
 Wed-0.029
Token was
Token ID373
Feature activation-9.70e-02

Pos logit contributions
Lost+0.196
Spider+0.195
mail+0.195
Ancient+0.195
 swarm+0.191
Neg logit contributions
 how-0.090
-0.074
 why-0.065
's-0.058
 just-0.056
Token only
Token ID691
Feature activation-1.24e-01

Pos logit contributions
mail+0.133
 faded+0.130
Factory+0.129
Spider+0.129
Ancient+0.127
Neg logit contributions
 how-0.090
 why-0.077
 what-0.077
 our-0.076
ric-0.074
Token fair
Token ID3148
Feature activation-1.64e-01

Pos logit contributions
Spider+0.181
 faded+0.176
mail+0.176
Bee+0.175
Lost+0.168
Neg logit contributions
 how-0.104
ric-0.087
 where-0.084
 why-0.083
 what-0.082
Token.
Token ID13
Feature activation+1.92e-02

Pos logit contributions
 faded+0.206
 melted+0.193
 tangled+0.193
 shone+0.192
 healed+0.185
Neg logit contributions
 Amen-0.163
iu-0.161
ric-0.154
ext-0.146
 interests-0.134
Token So
Token ID1406
Feature activation-1.21e-02

Pos logit contributions
iu+0.014
ric+0.013
 Amen+0.013
ext+0.012
 interests+0.011
Neg logit contributions
 faded-0.029
 shone-0.028
 tangled-0.028
 melted-0.028
chy-0.027
Token,
Token ID11
Feature activation-3.13e-02

Pos logit contributions
 faded+0.019
 tangled+0.018
chy+0.018
 shone+0.018
 melted+0.018
Neg logit contributions
iu-0.007
 Amen-0.007
ric-0.007
 how-0.007
ext-0.006
Token they
Token ID484
Feature activation-7.34e-02

Pos logit contributions
chy+0.051
 faded+0.051
 iv+0.050
 tangled+0.050
 shone+0.049
Neg logit contributions
iu-0.016
 how-0.016
ric-0.014
 Amen-0.014
 we-0.014
Token decided
Token ID3066
Feature activation-1.16e-01

Pos logit contributions
 tangled+0.096
chy+0.095
 stiff+0.095
 darker+0.094
Ancient+0.092
Neg logit contributions
 how-0.062
iu-0.056
 Amen-0.055
 spend-0.053
 why-0.049
Token to
Token ID284
Feature activation-8.52e-02

Pos logit contributions
 faded+0.173
 tangled+0.164
 melted+0.164
 shone+0.162
 healed+0.159
Neg logit contributions
iu-0.088
 Amen-0.086
ric-0.085
ext-0.077
 interests-0.071
Token let
Token ID1309
Feature activation-1.22e-01

Pos logit contributions
 faded+0.097
 melted+0.096
 healed+0.096
 shone+0.092
 disappeared+0.090
Neg logit contributions
iu-0.091
ext-0.085
 Amen-0.083
ric-0.083
icol-0.077
Token inside
Token ID2641
Feature activation-3.81e-02

Pos logit contributions
 tangled+0.023
 dark+0.022
 fast+0.022
 disappeared+0.022
 faded+0.021
Neg logit contributions
 Amen-0.055
ric-0.055
iu-0.051
 disl-0.046
 interests-0.046
Token.
Token ID13
Feature activation+9.80e-02

Pos logit contributions
 faded+0.058
 shone+0.055
 melted+0.055
 tangled+0.055
chy+0.055
Neg logit contributions
iu-0.023
 how-0.023
 Amen-0.023
ric-0.023
ext-0.022
Token As
Token ID1081
Feature activation+1.80e-02

Pos logit contributions
iu+0.078
ric+0.076
 Amen+0.076
ext+0.070
 interests+0.065
Neg logit contributions
 faded-0.140
 tangled-0.133
 shone-0.132
 melted-0.132
 healed-0.129
Token they
Token ID484
Feature activation+1.06e-02

Pos logit contributions
 we+0.007
iu+0.007
 how+0.006
 Wed+0.006
 our+0.006
Neg logit contributions
 iv-0.031
chy-0.031
mail-0.031
 flowed-0.030
 healed-0.030
Token stepped
Token ID10764
Feature activation-5.74e-02

Pos logit contributions
iu+0.010
 how+0.009
ext+0.009
ric+0.009
 spend+0.008
Neg logit contributions
chy-0.012
 dead-0.012
 faded-0.012
 wicked-0.012
 shadowy-0.011
Token over
Token ID625
Feature activation-7.16e-02

Pos logit contributions
 faded+0.069
 melted+0.064
 tangled+0.064
 shone+0.064
 healed+0.062
Neg logit contributions
iu-0.058
 Amen-0.058
ric-0.056
ext-0.054
 interests-0.050
Token the
Token ID262
Feature activation-4.25e-02

Pos logit contributions
chy+0.105
 faded+0.104
 healed+0.100
 shone+0.098
 flowed+0.098
Neg logit contributions
iu-0.048
 how-0.046
 Amen-0.044
ext-0.044
ric-0.044
Token threshold
Token ID11387
Feature activation-2.72e-02

Pos logit contributions
 faded+0.049
 flowed+0.046
 spoke+0.046
 attracted+0.046
 shone+0.045
Neg logit contributions
ric-0.041
iu-0.040
 Amen-0.040
icol-0.036
 Wed-0.035
Token they
Token ID484
Feature activation-1.40e-03

Pos logit contributions
mail+0.042
 tangled+0.041
 faded+0.041
chy+0.040
 healed+0.040
Neg logit contributions
iu-0.015
 how-0.015
ric-0.015
 Amen-0.013
 We-0.013
Token heard
Token ID2982
Feature activation-5.34e-02

Pos logit contributions
chini+0.002
 python+0.002
iky+0.002
adia+0.002
 poisonous+0.002
Neg logit contributions
 go-0.001
 each-0.001
 both-0.001
 come-0.001
 why-0.001
Token a
Token ID257
Feature activation-5.94e-02

Pos logit contributions
 tangled+0.082
chy+0.080
 faded+0.079
mail+0.077
 healed+0.077
Neg logit contributions
 how-0.041
 what-0.035
 why-0.033
iu-0.032
ric-0.031
Token buddy
Token ID24407
Feature activation-2.41e-02

Pos logit contributions
iu+0.003
ric+0.003
 Amen+0.003
ext+0.003
 interests+0.003
Neg logit contributions
 faded-0.002
 melted-0.002
 tangled-0.002
 shone-0.002
 healed-0.002
Token!
Token ID0
Feature activation+1.13e-01

Pos logit contributions
 faded+0.038
 tangled+0.037
 shone+0.036
 flowed+0.036
 melted+0.035
Neg logit contributions
 Amen-0.015
ric-0.015
iu-0.015
ext-0.013
 how-0.012
Token Do
Token ID2141
Feature activation+5.08e-02

Pos logit contributions
iu+0.118
ric+0.109
ext+0.108
 interests+0.102
pieces+0.099
Neg logit contributions
 White-0.127
 faded-0.125
 healed-0.121
 shone-0.121
 tangled-0.116
Token you
Token ID345
Feature activation+1.81e-02

Pos logit contributions
icol+0.088
 Amen+0.081
 stret+0.076
 impression+0.076
iu+0.075
Neg logit contributions
 bad-0.033
 short-0.032
 small-0.032
 other-0.031
 hurt-0.031
Token want
Token ID765
Feature activation+1.13e-02

Pos logit contributions
iu+0.019
ric+0.018
ext+0.018
 Amen+0.018
icol+0.017
Neg logit contributions
 faded-0.022
 melted-0.021
chy-0.021
 tangled-0.021
 healed-0.021
Token to
Token ID284
Feature activation+2.76e-02

Pos logit contributions
 Amen+0.006
iu+0.005
 how+0.005
ric+0.005
ext+0.005
Neg logit contributions
 faded-0.019
 tangled-0.018
chy-0.018
 flowed-0.018
 shone-0.018
Token come
Token ID1282
Feature activation+3.82e-03

Pos logit contributions
iu+0.031
ext+0.028
 Amen+0.028
ric+0.026
 interests+0.025
Neg logit contributions
 faded-0.033
 healed-0.032
 melted-0.032
 shone-0.031
 disappeared-0.031
Token take
Token ID1011
Feature activation-4.28e-02

Pos logit contributions
iu+0.005
ric+0.004
 Amen+0.004
ext+0.004
 interests+0.004
Neg logit contributions
 faded-0.004
 shone-0.004
 tangled-0.004
 healed-0.004
 melted-0.004
Token a
Token ID257
Feature activation-2.76e-02

Pos logit contributions
 faded+0.069
chy+0.065
 flowed+0.064
 shone+0.064
 tangled+0.064
Neg logit contributions
 Amen-0.029
iu-0.028
 how-0.025
ext-0.025
 you-0.024
Token closer
Token ID5699
Feature activation-5.36e-02

Pos logit contributions
 faded+0.034
 melted+0.033
 tangled+0.032
 shone+0.032
 healed+0.031
Neg logit contributions
iu-0.027
 Amen-0.026
ric-0.026
ext-0.025
 interests-0.023
Token look
Token ID804
Feature activation-5.23e-02

Pos logit contributions
 shone+0.056
 faded+0.056
 dark+0.055
 melted+0.054
 whispered+0.053
Neg logit contributions
iu-0.061
ext-0.059
 Amen-0.057
ric-0.057
 hey-0.054
Token and
Token ID290
Feature activation-4.31e-02

Pos logit contributions
 faded+0.034
 shone+0.031
chy+0.031
 melted+0.031
 tangled+0.030
Neg logit contributions
 Amen-0.037
iu-0.036
ric-0.035
 interests-0.034
ext-0.033
Token looked
Token ID3114
Feature activation-1.23e-01

Pos logit contributions
 tangled+0.041
chy+0.040
 faded+0.040
mail+0.040
 melted+0.039
Neg logit contributions
 how-0.050
 Amen-0.046
 what-0.046
iu-0.045
 why-0.045
Token up
Token ID510
Feature activation-1.39e-01

Pos logit contributions
 faded+0.124
 shone+0.118
 tangled+0.118
 melted+0.116
 disappeared+0.112
Neg logit contributions
iu-0.154
 Amen-0.151
ric-0.149
ext-0.141
 interests-0.137
Token.
Token ID13
Feature activation+2.36e-02

Pos logit contributions
 faded+0.150
 tangled+0.138
 dark+0.136
 melted+0.134
 shone+0.133
Neg logit contributions
ric-0.183
 Amen-0.181
iu-0.177
 interests-0.167
ext-0.162
Token There
Token ID1318
Feature activation-5.10e-02

Pos logit contributions
iu+0.022
 Amen+0.022
ric+0.021
ext+0.020
 interests+0.019
Neg logit contributions
 faded-0.030
 melted-0.029
 tangled-0.028
 shone-0.028
chy-0.027
Token was
Token ID373
Feature activation-1.09e-01

Pos logit contributions
 faded+0.071
 poisoned+0.070
 stiff+0.070
 tangled+0.069
 sore+0.069
Neg logit contributions
 how-0.033
ext-0.032
 what-0.031
-0.029
 why-0.028
Token a
Token ID257
Feature activation-1.25e-01

Pos logit contributions
 faded+0.148
 tangled+0.140
mail+0.140
 shone+0.137
 melted+0.137
Neg logit contributions
ric-0.085
 Amen-0.079
iu-0.079
 how-0.079
ext-0.078
Token fox
Token ID21831
Feature activation-1.44e-01

Pos logit contributions
 faded+0.115
 tangled+0.113
aked+0.112
 flowed+0.110
mail+0.108
Neg logit contributions
 Amen-0.154
ric-0.149
iu-0.138
icol-0.135
ext-0.126
Token!
Token ID0
Feature activation+1.23e-03

Pos logit contributions
 faded+0.158
 tangled+0.150
 shone+0.149
 melted+0.147
 disappeared+0.143
Neg logit contributions
iu-0.163
 Amen-0.159
ric-0.156
ext-0.149
 interests-0.145
Token The
Token ID383
Feature activation-9.67e-02

Pos logit contributions
 Amen+0.001
iu+0.001
 We+0.001
ric+0.001
 Wed+0.001
Neg logit contributions
 faded-0.002
 melted-0.002
chy-0.002
 shone-0.002
 tangled-0.002
Token fox
Token ID21831
Feature activation-1.52e-01

Pos logit contributions
 faded+0.093
 leaked+0.092
 flowed+0.089
 shone+0.087
 tram+0.087
Neg logit contributions
 Amen-0.109
 how-0.095
 you-0.092
 why-0.091
 what-0.090
Token out
Token ID503
Feature activation-2.24e-01

Pos logit contributions
 faded+0.287
Factory+0.270
 melted+0.269
 leaked+0.266
 flowed+0.264
Neg logit contributions
 how-0.161
ric-0.151
iu-0.149
ext-0.148
 Amen-0.145
Token and
Token ID290
Feature activation-1.45e-01

Pos logit contributions
mail+0.396
Factory+0.379
Ancient+0.376
Dead+0.374
chy+0.370
Neg logit contributions
 how-0.101
 where-0.071
 as-0.064
 what-0.064
iu-0.058
Token jumped
Token ID11687
Feature activation-2.33e-01

Pos logit contributions
Factory+0.153
umpy+0.145
iky+0.145
 iv+0.143
chy+0.142
Neg logit contributions
 how-0.162
iu-0.140
 what-0.137
 where-0.137
 we-0.136
Token onto
Token ID4291
Feature activation-2.38e-01

Pos logit contributions
 faded+0.333
 shone+0.319
mail+0.315
Dead+0.311
 smoked+0.310
Neg logit contributions
iu-0.165
ext-0.148
 how-0.147
 Amen-0.143
icol-0.142
Token Matt
Token ID4705
Feature activation-2.43e-01

Pos logit contributions
 faded+0.374
 shone+0.359
chy+0.355
 flowed+0.353
 tangled+0.351
Neg logit contributions
iu-0.159
 how-0.150
ric-0.143
 Wed-0.130
 Amen-0.129
Token's
Token ID338
Feature activation-1.62e-01

Pos logit contributions
 faded+0.385
 shone+0.366
 flowed+0.362
 melted+0.361
 tangled+0.359
Neg logit contributions
iu-0.152
ric-0.149
 Amen-0.136
ext-0.131
 interests-0.121
Token shoulders
Token ID12450
Feature activation-2.21e-01

Pos logit contributions
 faded+0.191
 shone+0.179
 tangled+0.178
 melted+0.174
 flowed+0.172
Neg logit contributions
iu-0.173
ric-0.166
 Amen-0.165
ext-0.156
 interests-0.149
Token.
Token ID13
Feature activation-8.96e-02

Pos logit contributions
 faded+0.321
 shone+0.306
 flowed+0.305
 melted+0.301
mail+0.300
Neg logit contributions
iu-0.153
ric-0.153
 Amen-0.140
 how-0.129
ext-0.128
Token The
Token ID383
Feature activation-1.93e-01

Pos logit contributions
 faded+0.124
 tangled+0.120
 melted+0.119
 shone+0.116
chy+0.115
Neg logit contributions
ric-0.077
iu-0.075
 Amen-0.072
ext-0.065
 Wed-0.064
Token monkey
Token ID21657
Feature activation-2.50e-01

Pos logit contributions
 faded+0.227
 shone+0.213
 flowed+0.210
mail+0.207
 tangled+0.200
Neg logit contributions
ric-0.192
iu-0.187
 Amen-0.177
 how-0.173
 interests-0.172
Token stayed
Token ID9658
Feature activation-2.77e-01

Pos logit contributions
 darker+0.350
eless+0.349
Factory+0.341
Ancient+0.339
 conditioning+0.329
Neg logit contributions
 how-0.218
-0.186
n-0.179
 then-0.175
 that-0.168
Token a
Token ID257
Feature activation-2.47e-01

Pos logit contributions
wy+0.279
chy+0.271
rito+0.270
iky+0.270
umpy+0.268
Neg logit contributions
 how-0.085
 what-0.068
 where-0.058
 that-0.056
 why-0.053
Token little
Token ID1310
Feature activation-2.44e-01

Pos logit contributions
aked+0.310
led+0.285
 flowed+0.279
 tangled+0.278
mail+0.277
Neg logit contributions
 Amen-0.243
ric-0.208
icol-0.187
iu-0.184
 please-0.181
Token boy
Token ID2933
Feature activation-2.65e-01

Pos logit contributions
mail+0.219
 flowed+0.214
 faded+0.206
 tangled+0.204
 contained+0.201
Neg logit contributions
 Amen-0.306
ric-0.278
 how-0.266
iu-0.260
 why-0.258
Token saw
Token ID2497
Feature activation-2.39e-01

Pos logit contributions
 tangled+0.396
 faded+0.388
 stiff+0.383
chy+0.379
 melted+0.372
Neg logit contributions
 Amen-0.170
iu-0.151
ric-0.150
 how-0.149
 why-0.135
Token it
Token ID340
Feature activation-2.72e-01

Pos logit contributions
 faded+0.418
chy+0.414
 tangled+0.405
 stiff+0.397
 sore+0.391
Neg logit contributions
 how-0.178
 what-0.146
 where-0.127
 why-0.121
ric-0.110
Token.
Token ID13
Feature activation-1.11e-01

Pos logit contributions
Ancient+0.499
Supp+0.491
Lost+0.488
Factory+0.486
mail+0.484
Neg logit contributions
 how-0.133
 where-0.107
 why-0.092
 what-0.091
ric-0.088
Token The
Token ID383
Feature activation-2.20e-01

Pos logit contributions
 faded+0.183
 tangled+0.179
 melted+0.172
 shone+0.172
 wil+0.169
Neg logit contributions
 Amen-0.068
 We-0.063
 how-0.063
ric-0.059
 Just-0.059
Token little
Token ID1310
Feature activation-2.56e-01

Pos logit contributions
 faded+0.294
 shone+0.268
 flowed+0.263
 disappeared+0.258
 spoke+0.256
Neg logit contributions
 Amen-0.200
 interests-0.165
 where-0.162
 how-0.160
iu-0.156
Token boy
Token ID2933
Feature activation-2.76e-01

Pos logit contributions
 faded+0.320
 flowed+0.294
chy+0.289
mail+0.284
 shone+0.279
Neg logit contributions
ric-0.219
 Amen-0.217
iu-0.212
 how-0.209
 interests-0.204
Token was
Token ID373
Feature activation-2.46e-01

Pos logit contributions
 tangled+0.398
 darker+0.395
eless+0.383
iest+0.381
Lost+0.380
Neg logit contributions
 how-0.242
 who-0.208
 why-0.198
 Amen-0.190
iu-0.185
Token only
Token ID691
Feature activation-2.74e-01

Pos logit contributions
 faded+0.261
mail+0.257
 shone+0.245
Factory+0.243
 flowed+0.243
Neg logit contributions
ric-0.280
 Amen-0.278
 how-0.256
ext-0.256
iu-0.249
Token played
Token ID2826
Feature activation-8.59e-02

Pos logit contributions
 lighter+0.084
 faded+0.082
 stiff+0.081
 swollen+0.080
 thinner+0.080
Neg logit contributions
 spend-0.057
 go-0.051
 how-0.050
iu-0.049
ap-0.048
Token in
Token ID287
Feature activation-6.65e-02

Pos logit contributions
 faded+0.090
 tangled+0.088
 shone+0.087
 melted+0.086
 flowed+0.085
Neg logit contributions
ric-0.100
iu-0.100
 Amen-0.100
ext-0.088
 interests-0.086
Token the
Token ID262
Feature activation-7.62e-02

Pos logit contributions
 tangled+0.039
 dark+0.037
 dead+0.034
 small+0.033
 melted+0.032
Neg logit contributions
iu-0.118
 Amen-0.117
ric-0.109
amb-0.107
icol-0.107
Token park
Token ID3952
Feature activation-8.10e-02

Pos logit contributions
 dark+0.052
 melted+0.048
 tangled+0.047
 shone+0.047
 dead+0.047
Neg logit contributions
iu-0.132
 Amen-0.129
ric-0.119
 interests-0.114
ext-0.111
Token and
Token ID290
Feature activation-2.15e-02

Pos logit contributions
 faded+0.110
 tangled+0.104
 melted+0.103
 shone+0.102
 healed+0.101
Neg logit contributions
iu-0.070
 Amen-0.068
ric-0.068
ext-0.064
 interests-0.060
Token laughed
Token ID13818
Feature activation-1.11e-01

Pos logit contributions
 faded+0.017
 shone+0.016
 tangled+0.016
 melted+0.016
 disappeared+0.015
Neg logit contributions
 Amen-0.031
iu-0.031
ric-0.031
ext-0.029
 interests-0.028
Token a
Token ID257
Feature activation-9.32e-02

Pos logit contributions
 faded+0.163
 tangled+0.157
 melted+0.155
 shone+0.151
 healed+0.150
Neg logit contributions
 Amen-0.080
iu-0.079
ext-0.077
 how-0.072
ric-0.071
Token lot
Token ID1256
Feature activation-7.31e-02

Pos logit contributions
 faded+0.085
 melted+0.079
 shone+0.078
 tangled+0.077
chy+0.074
Neg logit contributions
iu-0.124
ric-0.124
 Amen-0.121
ext-0.114
 interests-0.111
Token.
Token ID13
Feature activation+4.30e-02

Pos logit contributions
 faded+0.094
 tangled+0.088
 melted+0.088
 shone+0.088
 healed+0.085
Neg logit contributions
iu-0.069
 Amen-0.068
ric-0.067
ext-0.063
 interests-0.059
Token They
Token ID1119
Feature activation-2.28e-02

Pos logit contributions
 Amen+0.061
iu+0.060
ric+0.057
ext+0.052
earned+0.050
Neg logit contributions
 faded-0.037
 flowed-0.036
 tangled-0.036
 spilled-0.036
 shone-0.035
Token played
Token ID2826
Feature activation-1.11e-01

Pos logit contributions
 faded+0.029
 iv+0.029
 lighter+0.028
mail+0.028
 dies+0.028
Neg logit contributions
 how-0.019
 spend-0.019
 each-0.018
iu-0.016
ric-0.016
Token put
Token ID1234
Feature activation-2.35e-01

Pos logit contributions
 faded+0.266
 tangled+0.257
 healed+0.250
 melted+0.241
mail+0.240
Neg logit contributions
iu-0.291
ric-0.274
 Amen-0.268
 interests-0.263
 how-0.262
Token on
Token ID319
Feature activation-2.27e-01

Pos logit contributions
 faded+0.257
chy+0.230
 melted+0.226
 tangled+0.224
 healed+0.222
Neg logit contributions
iu-0.271
 Amen-0.260
ric-0.256
ext-0.248
 interests-0.237
Token their
Token ID511
Feature activation-2.12e-01

Pos logit contributions
 tangled+0.247
 melted+0.247
 faded+0.245
 shone+0.237
 flowed+0.231
Neg logit contributions
 Amen-0.271
ric-0.254
iu-0.246
ext-0.219
 interests-0.218
Token life
Token ID1204
Feature activation-2.63e-01

Pos logit contributions
 faded+0.234
 tangled+0.219
 shone+0.216
 melted+0.214
 healed+0.209
Neg logit contributions
iu-0.237
 Amen-0.231
ric-0.228
ext-0.220
 interests-0.210
Token jackets
Token ID35129
Feature activation-2.13e-01

Pos logit contributions
 faded+0.272
 shone+0.269
 tangled+0.268
 melted+0.263
 whispered+0.254
Neg logit contributions
 Amen-0.323
iu-0.310
ric-0.308
ext-0.284
 Wed-0.259
Token and
Token ID290
Feature activation-1.62e-01

Pos logit contributions
 faded+0.303
 melted+0.280
 tangled+0.278
 healed+0.276
 shone+0.273
Neg logit contributions
ric-0.166
iu-0.161
 Amen-0.157
ext-0.147
 how-0.143
Token w
Token ID266
Feature activation-2.25e-01

Pos logit contributions
 faded+0.143
 tangled+0.132
 melted+0.130
 shone+0.130
 healed+0.125
Neg logit contributions
iu-0.214
 Amen-0.211
ric-0.208
ext-0.200
 interests-0.193
Tokenaded
Token ID5286
Feature activation-2.83e-01

Pos logit contributions
 faded+0.256
 tangled+0.242
 shone+0.239
 melted+0.237
 healed+0.230
Neg logit contributions
iu-0.246
ric-0.237
 Amen-0.235
ext-0.230
 interests-0.218
Token out
Token ID503
Feature activation-2.47e-01

Pos logit contributions
 faded+0.238
 shone+0.236
 flowed+0.235
 tangled+0.233
 melted+0.230
Neg logit contributions
 Amen-0.415
ric-0.393
iu-0.377
ext-0.357
 interests-0.346
Token into
Token ID656
Feature activation-2.25e-01

Pos logit contributions
 faded+0.334
 melted+0.316
 shone+0.314
 healed+0.313
 tangled+0.312
Neg logit contributions
iu-0.203
ric-0.202
 Amen-0.192
ext-0.183
 how-0.179
Token the
Token ID262
Feature activation-2.30e-01

Pos logit contributions
 faded+0.242
 tangled+0.225
 melted+0.224
 shone+0.223
 healed+0.215
Neg logit contributions
iu-0.258
 Amen-0.254
ric-0.252
ext-0.238
 interests-0.227