TOP ACTIVATIONS
MAX = 2.106

worst year for jobs in almost two decades . " The
Irvine , California . Hide Caption 10 of 15 Photos :
a cerebral journalist played by Diane Ke aton , who proudly
do everything we can to find cool things to take their
constitution , and Kim Jong - il elected as chairman of
and not throw away their votes
No ar Ha O ved Ve Ha L omed , D
strictly prohibited without the express written permission of Money Morning .
Anyway , we had to throw a party ! Friends ,
) 6 . Rob Harley ( 152 ) (
, who have kept up their criticism of the election and
late 2011 . It took almost 2 years for Bitcoin prices
promote public health ," said Diane Brown , executive director of
un zel ?" I finally found my voice and gave her
53 p . m . PT with additional details and analysis
returned to Al cat raz nearly 50 years after the prison
much as he w ants to be able to
which I have kept for almost 30 years .
MORE and Hillary Clinton Hillary Diane Rodham Clinton Sanders : '
war and the contr aband trade .

MIN ACTIVATIONS
MIN = -2.083

ised on a vast m iscal culation about the origins of
where do we draw the line ? So they say ,
/ Office ) and enable Broad band Connection ( Co ax
a nightmare apartment broker on Broad City and once did cart
including House Majority Whip Steve Scal ise ( R - La
that today 's crash will inaug urate a bear market like
reached at j on . chest o @ gl obe .
HA W K AND AP AC HE T OC RE P
no internet connection , but Broad band Connection ( Co ax
The property now belongs to Cit ib ank .
( which has the Hyp ac ros aurus egg ) into
ag ression de t out cit oy en Imp é rial
h iking s holy gra il . In Chicago
with him . <|endoftext|> S AC RAM ENT O Thousands
:// t . co / l g S wh X Um
through a ridiculous and unpopular bill in just two weeks ?
, above all , an uncons c ion able gift to
flavor begins with dried apr ic ot , banana , and
might become worthless . The gap was filled by , among
Be that as it may , there seems to

INTERVAL 1.059 - 2.106
CONTAINS 0.482%

If we simply want to write an Own ed
start , after Ibrahim 's lawyers informed the DOJ that they
. S ., also dropped nearly 3 percent on the news
need it . RA ID 4 on tape .
The flyer shows a newspaper headline taken from the National Post

INTERVAL 0.011 - 1.059
CONTAINS 48.477%

Materials . ( Vis ited
Y alk ut Tal mud Torah , Genesis ,
the point where this sort of comparison works more as a
With the console method , you have to go
had known about his actions and had taken steps to get

INTERVAL -1.036 - 0.011
CONTAINS 50.461%

Dead line for Clubs to ask Players with "
, Ground Zero for social justice brain washing , the Social
s Diary . I think
like your looking at the image through a mesh ( usually
s nervous ness about the security forces

INTERVAL -2.083 - -1.036
CONTAINS 0.579%

, Robert Gibbs , the White House press secretary , made
deployed with maximum deployment in Roh t ak . Additional columns
room internet at the West in cut out ( as it
and touched off the political rift that paraly zes Palestinian politics
The statement by the FDA largely came as a surprise
Token worst
Token ID5290
Feature activation+6.95e-01

Pos logit contributions
 anonymously+1.330
orr+1.278
oux+1.233
 disse+1.200
church+1.199
Neg logit contributions
advertising-1.272
 Dresden-1.183
���-1.176
education-1.147
defense-1.145
Token year
Token ID614
Feature activation+5.85e-01

Pos logit contributions
advertising+4.228
ary+4.101
 Dresden+4.093
education+3.849
defense+3.813
Neg logit contributions
 anonymously-4.838
oux-4.714
orr-4.681
church-4.435
 disse-4.418
Token for
Token ID329
Feature activation+5.15e-01

Pos logit contributions
advertising+3.705
ary+3.540
 Dresden+3.514
education+3.330
defense+3.296
Neg logit contributions
 anonymously-4.005
orr-3.922
oux-3.879
 disse-3.741
church-3.687
Token jobs
Token ID3946
Feature activation+5.12e-01

Pos logit contributions
advertising+3.217
 Dresden+3.084
ary+2.993
education+2.933
defense+2.863
Neg logit contributions
 anonymously-3.598
orr-3.555
oux-3.483
 disse-3.337
church-3.302
Token in
Token ID287
Feature activation+1.23e+00

Pos logit contributions
advertising+3.096
 Dresden+2.894
ary+2.866
education+2.765
defense+2.702
Neg logit contributions
 anonymously-3.725
orr-3.671
oux-3.550
 disse-3.454
church-3.452
Token almost
Token ID2048
Feature activation+2.11e+00

Pos logit contributions
 Dresden+6.133
advertising+6.100
ary+5.841
education+5.635
defense+5.371
Neg logit contributions
 anonymously-9.098
orr-9.029
oux-8.865
 disse-8.557
church-8.444
Token two
Token ID734
Feature activation+1.66e+00

Pos logit contributions
 Dresden+9.044
advertising+8.274
ary+8.089
education+7.540
 Morales+7.216
Neg logit contributions
orr-15.060
oux-14.804
ococ-14.524
 anonymously-14.290
 disse-14.133
Token decades
Token ID4647
Feature activation+3.24e-01

Pos logit contributions
advertising+7.019
 Dresden+6.880
ary+6.799
education+6.159
 Morales+5.741
Neg logit contributions
orr-12.742
 anonymously-12.618
oux-12.612
kef-12.039
church-12.030
Token.
Token ID13
Feature activation+5.22e-04

Pos logit contributions
advertising+2.080
 Dresden+1.941
ary+1.900
education+1.863
defense+1.848
Neg logit contributions
 anonymously-2.292
orr-2.229
oux-2.168
 disse-2.112
church-2.097
Token "
Token ID366
Feature activation+6.32e-03

Pos logit contributions
advertising+0.004
 Dresden+0.003
���+0.003
education+0.003
defense+0.003
Neg logit contributions
 anonymously-0.004
orr-0.003
oux-0.003
 disse-0.003
church-0.003
TokenThe
Token ID464
Feature activation-1.94e-01

Pos logit contributions
advertising+0.042
 Dresden+0.038
���+0.037
education+0.037
defense+0.037
Neg logit contributions
 anonymously-0.045
orr-0.043
oux-0.042
 disse-0.041
church-0.041
Token Irvine
Token ID37277
Feature activation-8.61e-02

Pos logit contributions
advertising+1.069
 Dresden+0.972
ary+0.911
���+0.910
education+0.906
Neg logit contributions
 anonymously-2.151
orr-2.103
oux-2.043
 disse-2.010
church-1.997
Token,
Token ID11
Feature activation+2.99e-01

Pos logit contributions
 anonymously+0.486
orr+0.463
oux+0.443
 disse+0.427
church+0.427
Neg logit contributions
advertising-0.689
 Dresden-0.649
���-0.646
education-0.633
defense-0.631
Token California
Token ID3442
Feature activation-1.58e-01

Pos logit contributions
advertising+1.312
 Dresden+1.173
education+1.133
���+1.130
defense+1.121
Neg logit contributions
 anonymously-2.704
orr-2.668
oux-2.584
 disse-2.551
church-2.548
Token.
Token ID13
Feature activation+2.63e-01

Pos logit contributions
 anonymously+0.980
orr+0.929
oux+0.900
church+0.869
 disse+0.853
Neg logit contributions
advertising-1.170
���-1.113
 Dresden-1.107
education-1.071
defense-1.068
Token Hide
Token ID10415
Feature activation+8.80e-01

Pos logit contributions
advertising+1.785
 Dresden+1.669
���+1.606
ary+1.605
education+1.599
Neg logit contributions
 anonymously-1.796
orr-1.730
oux-1.657
 disse-1.650
church-1.618
Token Caption
Token ID11260
Feature activation+2.07e+00

Pos logit contributions
advertising+5.493
 Dresden+5.209
ary+5.153
education+4.847
 Morales+4.721
Neg logit contributions
 anonymously-6.157
 disse-6.099
orr-5.886
oux-5.684
church-5.609
Token 10
Token ID838
Feature activation-1.87e-01

Pos logit contributions
advertising+5.980
 Dresden+5.245
ary+5.239
 Morales+4.828
 Helpful+4.444
Neg logit contributions
 mosqu-19.699
 disse-19.264
 cryst-19.120
oux-18.912
church-18.796
Token of
Token ID286
Feature activation+3.85e-01

Pos logit contributions
 anonymously+1.313
orr+1.262
oux+1.218
church+1.184
 disse+1.170
Neg logit contributions
advertising-1.234
���-1.190
 Dresden-1.151
education-1.116
defense-1.114
Token 15
Token ID1315
Feature activation+1.70e-02

Pos logit contributions
advertising+1.884
 Dresden+1.747
ary+1.683
education+1.624
defense+1.616
Neg logit contributions
 anonymously-3.328
orr-3.245
 disse-3.170
oux-3.138
church-3.086
Token Photos
Token ID9434
Feature activation+2.20e-01

Pos logit contributions
advertising+0.142
 Dresden+0.134
ary+0.130
education+0.130
defense+0.130
Neg logit contributions
 anonymously-0.091
orr-0.086
oux-0.082
 disse-0.081
church-0.079
Token:
Token ID25
Feature activation+6.52e-01

Pos logit contributions
advertising+1.756
ary+1.632
 Dresden+1.629
education+1.587
defense+1.577
Neg logit contributions
 anonymously-1.270
orr-1.206
 disse-1.186
oux-1.158
church-1.120
Token a
Token ID257
Feature activation+1.92e-01

Pos logit contributions
advertising+2.012
 Dresden+1.924
ary+1.854
���+1.829
education+1.819
Neg logit contributions
 anonymously-1.781
orr-1.750
oux-1.682
church-1.639
 disse-1.600
Token cerebral
Token ID31169
Feature activation-4.14e-01

Pos logit contributions
advertising+1.318
 Dresden+1.254
���+1.211
ary+1.204
education+1.197
Neg logit contributions
 anonymously-1.275
orr-1.236
oux-1.189
church-1.157
 disse-1.143
Token journalist
Token ID10099
Feature activation+4.09e-01

Pos logit contributions
 anonymously+2.627
orr+2.494
oux+2.402
 disse+2.375
church+2.324
Neg logit contributions
advertising-3.024
���-2.840
 Dresden-2.783
defense-2.738
education-2.732
Token played
Token ID2826
Feature activation+4.52e-01

Pos logit contributions
advertising+2.878
 Dresden+2.775
ary+2.693
education+2.648
���+2.647
Neg logit contributions
 anonymously-2.541
orr-2.508
oux-2.394
church-2.322
 disse-2.285
Token by
Token ID416
Feature activation+1.12e-01

Pos logit contributions
advertising+2.511
 Dresden+2.376
���+2.267
ary+2.267
education+2.233
Neg logit contributions
 anonymously-3.510
orr-3.466
oux-3.327
church-3.278
 disse-3.258
Token Diane
Token ID26542
Feature activation+2.05e+00

Pos logit contributions
advertising+0.731
 Dresden+0.684
���+0.672
education+0.659
ary+0.657
Neg logit contributions
 anonymously-0.792
orr-0.764
oux-0.737
church-0.718
 disse-0.717
Token Ke
Token ID3873
Feature activation+1.49e+00

Pos logit contributions
 Dresden+9.732
advertising+9.693
ary+9.114
 Morales+9.068
lance+8.554
Neg logit contributions
 anonymously-14.878
orr-14.044
 disse-13.670
church-13.494
scribe-13.465
Tokenaton
Token ID13951
Feature activation+8.26e-01

Pos logit contributions
advertising+7.446
ary+6.806
education+6.244
peria+6.153
amily+6.089
Neg logit contributions
 anonymously-12.395
 disse-11.665
orr-10.959
church-10.531
kef-10.522
Token,
Token ID11
Feature activation+3.52e-01

Pos logit contributions
advertising+5.261
 Dresden+4.954
ary+4.904
education+4.692
ebted+4.677
Neg logit contributions
 anonymously-5.528
orr-5.337
oux-5.126
church-5.092
 disse-5.012
Token who
Token ID508
Feature activation+4.88e-01

Pos logit contributions
advertising+2.482
 Dresden+2.356
ary+2.275
���+2.268
education+2.259
Neg logit contributions
 anonymously-2.244
orr-2.198
oux-2.109
church-2.052
 disse-2.030
Token proudly
Token ID23017
Feature activation+3.30e-01

Pos logit contributions
advertising+3.348
 Dresden+3.149
ary+3.101
���+3.044
education+3.040
Neg logit contributions
 anonymously-3.096
orr-3.029
oux-2.932
church-2.892
 disse-2.814
Token do
Token ID466
Feature activation+1.05e+00

Pos logit contributions
advertising+4.956
ary+4.888
 Dresden+4.884
education+4.685
defense+4.545
Neg logit contributions
orr-4.960
 anonymously-4.902
oux-4.901
church-4.851
 disse-4.740
Token everything
Token ID2279
Feature activation+1.31e+00

Pos logit contributions
advertising+5.002
 Dresden+4.798
ary+4.495
education+4.489
���+4.312
Neg logit contributions
 anonymously-8.488
orr-8.400
oux-7.993
church-7.939
 disse-7.848
Token we
Token ID356
Feature activation+1.34e+00

Pos logit contributions
advertising+6.592
 Dresden+6.178
ary+6.094
education+5.696
defense+5.539
Neg logit contributions
 anonymously-9.670
orr-9.667
oux-9.487
church-9.456
 disse-9.347
Token can
Token ID460
Feature activation+7.87e-01

Pos logit contributions
advertising+7.324
education+6.769
ary+6.673
 Dresden+6.671
defense+6.487
Neg logit contributions
oux-9.046
orr-9.042
church-9.039
 anonymously-8.796
 disse-8.644
Token to
Token ID284
Feature activation+9.62e-01

Pos logit contributions
advertising+4.258
 Dresden+4.024
ary+3.933
education+3.845
defense+3.735
Neg logit contributions
 anonymously-5.882
orr-5.881
oux-5.736
 disse-5.592
church-5.582
Token find
Token ID1064
Feature activation+2.00e+00

Pos logit contributions
advertising+5.768
 Dresden+5.637
ary+5.561
education+5.410
defense+5.296
Neg logit contributions
orr-6.126
oux-6.040
 anonymously-6.006
church-5.989
 disse-5.784
Token cool
Token ID3608
Feature activation+3.66e-02

Pos logit contributions
 Dresden+9.701
advertising+9.537
ary+9.382
 contestant+8.833
education+8.722
Neg logit contributions
oux-12.683
orr-12.600
 anonymously-12.506
church-12.354
kef-12.173
Token things
Token ID1243
Feature activation+2.15e-01

Pos logit contributions
advertising+0.258
 Dresden+0.242
���+0.234
education+0.233
ary+0.233
Neg logit contributions
 anonymously-0.242
orr-0.232
oux-0.225
church-0.218
 disse-0.217
Token to
Token ID284
Feature activation+9.60e-02

Pos logit contributions
advertising+1.411
 Dresden+1.323
ary+1.271
���+1.260
education+1.258
Neg logit contributions
 anonymously-1.485
orr-1.455
oux-1.425
 disse-1.377
church-1.370
Token take
Token ID1011
Feature activation-5.65e-01

Pos logit contributions
advertising+0.667
 Dresden+0.631
education+0.611
ary+0.610
���+0.607
Neg logit contributions
 anonymously-0.619
orr-0.600
oux-0.587
church-0.574
 disse-0.570
Token their
Token ID511
Feature activation+2.18e-01

Pos logit contributions
 anonymously+3.634
orr+3.454
oux+3.301
church+3.205
 disse+3.173
Neg logit contributions
advertising-4.076
���-3.829
 Dresden-3.766
peria-3.682
education-3.682
Token constitution
Token ID11784
Feature activation+5.00e-01

Pos logit contributions
advertising+3.710
 Dresden+3.512
ary+3.455
education+3.447
defense+3.393
Neg logit contributions
 anonymously-3.945
orr-3.881
oux-3.739
 disse-3.566
church-3.555
Token,
Token ID11
Feature activation-4.56e-01

Pos logit contributions
advertising+3.375
 Dresden+3.187
ary+3.167
education+3.095
defense+3.074
Neg logit contributions
 anonymously-3.267
orr-3.211
oux-3.088
 disse-3.017
church-3.001
Token and
Token ID290
Feature activation-1.05e-01

Pos logit contributions
 anonymously+2.965
orr+2.798
oux+2.692
church+2.651
 disse+2.630
Neg logit contributions
advertising-3.232
���-3.034
 Dresden-3.004
peria-2.940
defense-2.900
Token Kim
Token ID6502
Feature activation+1.02e+00

Pos logit contributions
 anonymously+0.727
orr+0.699
oux+0.675
 disse+0.655
church+0.655
Neg logit contributions
advertising-0.708
 Dresden-0.660
���-0.654
education-0.640
defense-0.639
Token Jong
Token ID17960
Feature activation+1.56e+00

Pos logit contributions
advertising+5.560
 Dresden+5.449
ary+5.375
���+5.240
defense+5.192
Neg logit contributions
 anonymously-7.350
oux-7.063
orr-7.060
 disse-6.854
church-6.798
Token-
Token ID12
Feature activation+1.95e+00

Pos logit contributions
advertising+7.506
 Dresden+7.275
education+6.899
ary+6.859
defense+6.854
Neg logit contributions
 anonymously-11.371
orr-11.040
oux-10.861
church-10.681
 disse-10.551
Tokenil
Token ID346
Feature activation+4.29e-01

Pos logit contributions
advertising+8.781
education+8.504
defense+8.206
ary+7.765
amily+7.549
Neg logit contributions
 anonymously-14.332
 disse-14.042
oux-13.430
orr-13.398
church-12.912
Token elected
Token ID7018
Feature activation-6.20e-01

Pos logit contributions
advertising+2.932
 Dresden+2.780
ary+2.757
���+2.703
education+2.690
Neg logit contributions
 anonymously-2.793
orr-2.696
oux-2.578
 disse-2.560
church-2.533
Token as
Token ID355
Feature activation+5.40e-01

Pos logit contributions
 anonymously+4.520
orr+4.229
oux+4.075
 disse+4.026
church+4.018
Neg logit contributions
advertising-3.945
���-3.654
 Dresden-3.524
peria-3.504
ebted-3.460
Token chairman
Token ID8900
Feature activation+6.20e-01

Pos logit contributions
advertising+3.780
 Dresden+3.695
defense+3.625
education+3.618
ary+3.601
Neg logit contributions
 anonymously-3.340
orr-3.288
oux-3.166
 disse-3.024
church-2.984
Token of
Token ID286
Feature activation+1.07e-01

Pos logit contributions
advertising+3.793
 Dresden+3.643
ary+3.513
defense+3.495
education+3.486
Neg logit contributions
 anonymously-4.423
orr-4.320
oux-4.182
 disse-4.060
church-4.027
Token
Token ID251
Feature activation-2.07e-01

Pos logit contributions
 anonymously+2.116
orr+2.048
oux+1.971
church+1.932
 disse+1.911
Neg logit contributions
advertising-1.671
 Dresden-1.546
���-1.533
peria-1.501
defense-1.493
Token and
Token ID290
Feature activation-1.84e-01

Pos logit contributions
 anonymously+1.377
orr+1.319
oux+1.270
church+1.234
 disse+1.230
Neg logit contributions
advertising-1.448
���-1.355
 Dresden-1.352
defense-1.308
peria-1.306
Token not
Token ID407
Feature activation-3.24e-02

Pos logit contributions
 anonymously+1.217
orr+1.161
oux+1.117
 disse+1.088
church+1.084
Neg logit contributions
advertising-1.294
 Dresden-1.205
���-1.204
education-1.168
defense-1.167
Token
Token ID564
Feature activation-1.11e-01

Pos logit contributions
 anonymously+0.211
orr+0.203
oux+0.196
church+0.190
 disse+0.190
Neg logit contributions
advertising-0.230
 Dresden-0.215
���-0.213
education-0.209
defense-0.208
Token
Token ID250
Feature activation+7.39e-02

Pos logit contributions
 anonymously+0.894
orr+0.865
oux+0.837
church+0.819
 disse+0.817
Neg logit contributions
advertising-0.620
 Dresden-0.569
���-0.563
education-0.546
peria-0.545
Tokenthrow
Token ID16939
Feature activation+1.86e+00

Pos logit contributions
advertising+0.496
 Dresden+0.459
���+0.454
education+0.449
defense+0.448
Neg logit contributions
 anonymously-0.514
orr-0.493
oux-0.476
 disse-0.465
church-0.461
Token away
Token ID1497
Feature activation+5.50e-01

Pos logit contributions
ary+11.533
advertising+11.497
 Dresden+10.845
education+10.422
defense+10.182
Neg logit contributions
 anonymously-10.946
oux-10.576
 disse-10.327
orr-10.039
church-9.895
Token
Token ID447
Feature activation+1.13e-02

Pos logit contributions
advertising+3.500
 Dresden+3.325
ary+3.224
education+3.223
���+3.211
Neg logit contributions
 anonymously-3.854
orr-3.700
oux-3.609
 disse-3.539
church-3.526
Token
Token ID251
Feature activation+3.30e-01

Pos logit contributions
advertising+0.060
 Dresden+0.055
���+0.055
education+0.053
defense+0.053
Neg logit contributions
 anonymously-0.094
orr-0.090
oux-0.088
 disse-0.086
church-0.086
Token their
Token ID511
Feature activation+7.98e-01

Pos logit contributions
advertising+2.226
 Dresden+2.087
ary+2.047
education+2.047
���+2.018
Neg logit contributions
 anonymously-2.242
orr-2.145
oux-2.089
 disse-2.042
church-2.016
Token votes
Token ID5690
Feature activation+3.91e-01

Pos logit contributions
advertising+5.181
 Dresden+5.054
ary+4.896
education+4.824
defense+4.715
Neg logit contributions
 anonymously-5.147
orr-5.109
oux-4.988
church-4.834
ucci-4.718
TokenNo
Token ID2949
Feature activation+4.46e-01

Pos logit contributions
advertising+4.276
 Dresden+3.884
ary+3.879
defense+3.867
peria+3.755
Neg logit contributions
 anonymously-5.626
orr-5.310
 disse-5.244
oux-5.204
church-5.129
Tokenar
Token ID283
Feature activation+7.64e-02

Pos logit contributions
advertising+2.847
 Dresden+2.685
defense+2.635
ary+2.633
���+2.624
Neg logit contributions
 anonymously-3.126
orr-2.962
oux-2.911
 disse-2.852
church-2.843
Token Ha
Token ID9398
Feature activation+9.87e-01

Pos logit contributions
advertising+0.546
 Dresden+0.515
���+0.509
education+0.499
defense+0.498
Neg logit contributions
 anonymously-0.492
orr-0.472
oux-0.453
 disse-0.443
church-0.442
TokenO
Token ID46
Feature activation+7.76e-01

Pos logit contributions
advertising+6.035
defense+5.555
 Dresden+5.554
peria+5.551
ary+5.545
Neg logit contributions
 anonymously-6.812
orr-6.376
oux-6.340
 disse-6.296
church-6.220
Tokenved
Token ID1079
Feature activation+6.60e-01

Pos logit contributions
advertising+5.369
defense+4.984
education+4.896
 Dresden+4.868
ary+4.861
Neg logit contributions
 anonymously-4.974
orr-4.558
 disse-4.486
oux-4.384
church-4.240
Token Ve
Token ID8016
Feature activation+1.86e+00

Pos logit contributions
advertising+4.061
ary+3.845
 Dresden+3.813
education+3.757
defense+3.731
Neg logit contributions
 anonymously-4.617
orr-4.435
oux-4.277
 disse-4.254
church-4.248
TokenHa
Token ID23303
Feature activation+1.03e+00

Pos logit contributions
advertising+11.446
ary+11.169
amily+11.013
peria+10.898
defense+10.872
Neg logit contributions
 anonymously-10.235
oux-9.253
church-9.114
 disse-8.876
orr-8.846
TokenL
Token ID43
Feature activation+1.53e-01

Pos logit contributions
advertising+6.538
ary+6.286
 Dresden+6.253
defense+6.161
���+6.132
Neg logit contributions
 anonymously-6.503
orr-6.223
oux-6.083
church-6.041
 disse-6.004
Tokenomed
Token ID12657
Feature activation+5.92e-01

Pos logit contributions
advertising+1.051
 Dresden+0.983
���+0.974
education+0.960
ary+0.958
Neg logit contributions
 anonymously-1.021
orr-0.975
oux-0.940
 disse-0.920
church-0.918
Token,
Token ID11
Feature activation-4.91e-01

Pos logit contributions
advertising+4.011
 Dresden+3.856
ary+3.804
���+3.736
defense+3.732
Neg logit contributions
 anonymously-3.772
orr-3.617
oux-3.463
church-3.456
 disse-3.432
Token D
Token ID360
Feature activation+1.72e-01

Pos logit contributions
 anonymously+3.245
orr+3.055
oux+2.941
 disse+2.841
church+2.834
Neg logit contributions
advertising-3.448
���-3.200
 Dresden-3.155
peria-3.135
education-3.109
Token strictly
Token ID14084
Feature activation-7.69e-01

Pos logit contributions
 anonymously+2.947
orr+2.775
oux+2.643
 disse+2.602
church+2.587
Neg logit contributions
advertising-3.161
 Dresden-2.941
���-2.901
education-2.895
defense-2.884
Token prohibited
Token ID12244
Feature activation+4.66e-02

Pos logit contributions
 anonymously+4.515
orr+4.313
oux+3.964
church+3.923
 disse+3.892
Neg logit contributions
advertising-5.864
 Dresden-5.477
education-5.432
peria-5.419
���-5.402
Token without
Token ID1231
Feature activation+1.20e+00

Pos logit contributions
advertising+0.339
 Dresden+0.317
���+0.313
education+0.307
ary+0.307
Neg logit contributions
 anonymously-0.295
orr-0.285
oux-0.275
church-0.267
 disse-0.265
Token the
Token ID262
Feature activation+1.22e+00

Pos logit contributions
advertising+6.711
���+6.484
ary+6.477
 Dresden+6.444
education+6.110
Neg logit contributions
 anonymously-8.145
orr-8.139
oux-7.947
church-7.786
kef-7.610
Token express
Token ID4911
Feature activation+8.34e-01

Pos logit contributions
advertising+6.666
 Dresden+6.278
ary+6.143
���+6.036
education+5.961
Neg logit contributions
 anonymously-8.707
orr-8.470
oux-8.079
church-8.065
kef-8.027
Token written
Token ID3194
Feature activation+1.85e+00

Pos logit contributions
advertising+4.790
ary+4.531
 Dresden+4.526
education+4.330
defense+4.269
Neg logit contributions
 anonymously-6.076
orr-5.913
oux-5.646
church-5.519
ucci-5.476
Token permission
Token ID7170
Feature activation+3.76e-01

Pos logit contributions
ary+6.412
 Dresden+6.406
advertising+6.330
defense+5.577
 Morales+5.542
Neg logit contributions
orr-14.988
 anonymously-14.589
kef-14.309
ucci-14.306
oux-14.173
Token of
Token ID286
Feature activation+4.63e-01

Pos logit contributions
advertising+2.357
 Dresden+2.182
���+2.169
ary+2.112
education+2.107
Neg logit contributions
 anonymously-2.750
orr-2.657
oux-2.576
 disse-2.513
church-2.510
Token Money
Token ID12911
Feature activation-4.40e-01

Pos logit contributions
advertising+2.930
 Dresden+2.734
���+2.644
ary+2.631
education+2.609
Neg logit contributions
 anonymously-3.245
orr-3.167
oux-3.099
church-2.999
 disse-2.975
Token Morning
Token ID14410
Feature activation+5.89e-02

Pos logit contributions
 anonymously+2.502
orr+2.379
oux+2.269
church+2.202
 disse+2.159
Neg logit contributions
advertising-3.481
���-3.364
 Dresden-3.262
defense-3.199
peria-3.193
Token.
Token ID13
Feature activation+2.78e-01

Pos logit contributions
advertising+0.423
 Dresden+0.396
���+0.390
education+0.384
ary+0.383
Neg logit contributions
 anonymously-0.379
orr-0.366
oux-0.352
 disse-0.342
church-0.341
Token Anyway
Token ID21836
Feature activation-1.79e-01

Pos logit contributions
 anonymously+1.504
orr+1.446
oux+1.392
church+1.349
 disse+1.346
Neg logit contributions
advertising-1.547
���-1.456
 Dresden-1.442
education-1.407
defense-1.403
Token,
Token ID11
Feature activation+2.26e-01

Pos logit contributions
 anonymously+1.067
orr+1.018
oux+0.976
 disse+0.945
church+0.943
Neg logit contributions
advertising-1.378
 Dresden-1.294
���-1.289
education-1.261
defense-1.258
Token we
Token ID356
Feature activation+4.22e-01

Pos logit contributions
advertising+1.489
 Dresden+1.396
defense+1.341
ary+1.336
���+1.336
Neg logit contributions
 anonymously-1.559
orr-1.517
oux-1.459
church-1.436
 disse-1.420
Token had
Token ID550
Feature activation+2.26e-01

Pos logit contributions
advertising+2.883
 Dresden+2.774
ary+2.698
���+2.653
education+2.651
Neg logit contributions
 anonymously-2.672
orr-2.631
oux-2.531
church-2.501
 disse-2.419
Token to
Token ID284
Feature activation+5.36e-01

Pos logit contributions
advertising+1.358
 Dresden+1.275
���+1.248
ary+1.226
education+1.221
Neg logit contributions
 anonymously-1.683
orr-1.630
oux-1.570
church-1.546
 disse-1.540
Token throw
Token ID3714
Feature activation+1.85e+00

Pos logit contributions
advertising+3.608
 Dresden+3.571
ary+3.441
defense+3.432
education+3.377
Neg logit contributions
orr-3.263
 anonymously-3.255
church-3.143
oux-3.137
 disse-3.036
Token a
Token ID257
Feature activation+5.49e-01

Pos logit contributions
 Dresden+9.978
advertising+9.963
ary+9.846
 Morales+9.343
defense+9.091
Neg logit contributions
 anonymously-11.216
oux-11.048
orr-11.023
church-10.917
ococ-10.687
Token party
Token ID2151
Feature activation-1.56e-01

Pos logit contributions
advertising+3.758
 Dresden+3.702
ary+3.557
defense+3.489
���+3.453
Neg logit contributions
 anonymously-3.493
orr-3.415
oux-3.305
church-3.181
 disse-3.130
Token!
Token ID0
Feature activation+1.24e-01

Pos logit contributions
 anonymously+1.020
orr+0.973
oux+0.935
 disse+0.909
church+0.907
Neg logit contributions
advertising-1.116
���-1.044
 Dresden-1.041
education-1.013
defense-1.011
Token Friends
Token ID14213
Feature activation+1.29e-01

Pos logit contributions
advertising+0.834
 Dresden+0.784
���+0.754
education+0.751
ary+0.750
Neg logit contributions
 anonymously-0.854
orr-0.820
oux-0.792
 disse-0.775
church-0.771
Token,
Token ID11
Feature activation-3.20e-01

Pos logit contributions
advertising+0.922
 Dresden+0.873
���+0.851
ary+0.845
education+0.842
Neg logit contributions
 anonymously-0.819
orr-0.792
oux-0.764
 disse-0.737
church-0.737
Token)
Token ID8
Feature activation-3.60e-01

Pos logit contributions
 anonymously+2.220
orr+2.144
oux+2.067
church+2.009
 disse+1.971
Neg logit contributions
advertising-2.120
���-2.075
 Dresden-1.986
education-1.925
peria-1.922
Token
Token ID198
Feature activation-1.80e-01

Pos logit contributions
 anonymously+2.170
orr+2.067
oux+1.989
church+1.921
 disse+1.902
Neg logit contributions
advertising-2.729
���-2.569
 Dresden-2.536
education-2.499
peria-2.490
Token
Token ID198
Feature activation-3.22e-01

Pos logit contributions
 anonymously+1.120
orr+1.071
oux+1.028
church+0.996
 disse+0.994
Neg logit contributions
advertising-1.338
���-1.256
 Dresden-1.255
education-1.220
defense-1.218
Token6
Token ID21
Feature activation-1.79e-01

Pos logit contributions
 anonymously+2.262
orr+2.188
oux+2.112
church+2.064
 disse+2.025
Neg logit contributions
advertising-2.073
���-2.004
 Dresden-1.968
education-1.886
peria-1.879
Token.
Token ID13
Feature activation+2.25e-01

Pos logit contributions
 anonymously+1.135
orr+1.090
oux+1.046
church+1.013
 disse+1.001
Neg logit contributions
advertising-1.297
���-1.249
 Dresden-1.212
education-1.188
defense-1.185
Token Rob
Token ID3851
Feature activation+1.85e+00

Pos logit contributions
advertising+1.512
 Dresden+1.416
ary+1.366
���+1.359
education+1.354
Neg logit contributions
 anonymously-1.565
orr-1.495
oux-1.446
 disse-1.437
church-1.403
Token Harley
Token ID27440
Feature activation-9.76e-01

Pos logit contributions
advertising+6.932
ary+6.602
lance+6.479
 Dresden+6.399
���+5.956
Neg logit contributions
 anonymously-14.895
 disse-14.288
orr-14.279
oux-13.616
church-13.360
Token (
Token ID357
Feature activation-3.15e-01

Pos logit contributions
 anonymously+6.572
oux+6.211
orr+6.155
church+5.830
 disse+5.747
Neg logit contributions
advertising-6.324
���-5.957
peria-5.846
education-5.831
 Dresden-5.666
Token152
Token ID17827
Feature activation+2.22e-01

Pos logit contributions
 anonymously+1.869
orr+1.801
oux+1.721
church+1.680
 disse+1.633
Neg logit contributions
advertising-2.384
���-2.315
 Dresden-2.264
peria-2.190
education-2.190
Token)
Token ID8
Feature activation-2.23e-01

Pos logit contributions
advertising+1.329
 Dresden+1.211
ary+1.194
defense+1.170
education+1.167
Neg logit contributions
 anonymously-1.699
orr-1.626
 disse-1.601
oux-1.584
church-1.542
Token (
Token ID357
Feature activation-7.58e-02

Pos logit contributions
 anonymously+1.778
orr+1.717
oux+1.666
church+1.624
 disse+1.621
Neg logit contributions
advertising-1.260
���-1.163
 Dresden-1.155
education-1.116
defense-1.113
Token,
Token ID11
Feature activation+7.25e-02

Pos logit contributions
advertising+3.778
 Dresden+3.582
ary+3.497
education+3.470
defense+3.465
Neg logit contributions
 anonymously-3.405
orr-3.330
oux-3.282
church-3.186
 disse-3.175
Token who
Token ID508
Feature activation+4.15e-01

Pos logit contributions
advertising+0.519
 Dresden+0.490
education+0.478
defense+0.477
���+0.477
Neg logit contributions
 anonymously-0.461
orr-0.447
oux-0.433
church-0.418
 disse-0.416
Token have
Token ID423
Feature activation+1.84e-01

Pos logit contributions
advertising+2.637
 Dresden+2.576
ary+2.492
���+2.456
defense+2.454
Neg logit contributions
 anonymously-2.763
orr-2.761
oux-2.720
church-2.644
 disse-2.559
Token kept
Token ID4030
Feature activation+1.01e+00

Pos logit contributions
advertising+1.211
 Dresden+1.163
ary+1.128
���+1.121
defense+1.111
Neg logit contributions
 anonymously-1.223
orr-1.218
oux-1.186
church-1.150
 disse-1.117
Token up
Token ID510
Feature activation+6.68e-01

Pos logit contributions
 Dresden+6.436
advertising+6.273
ary+6.177
defense+5.910
education+5.903
Neg logit contributions
orr-6.353
oux-6.197
 anonymously-6.127
church-5.853
nect-5.779
Token their
Token ID511
Feature activation+1.83e+00

Pos logit contributions
advertising+4.302
 Dresden+4.149
ary+4.048
education+3.995
defense+3.978
Neg logit contributions
 anonymously-4.410
orr-4.319
oux-4.168
church-4.004
nect-3.993
Token criticism
Token ID7734
Feature activation+8.13e-02

Pos logit contributions
 Dresden+10.104
ary+9.352
advertising+9.149
defense+8.902
education+8.819
Neg logit contributions
orr-11.546
nect-11.200
oux-11.136
church-10.741
 anonymously-10.704
Token of
Token ID286
Feature activation-8.27e-01

Pos logit contributions
advertising+0.541
 Dresden+0.505
���+0.497
ary+0.490
defense+0.490
Neg logit contributions
 anonymously-0.562
orr-0.543
oux-0.527
 disse-0.511
church-0.511
Token the
Token ID262
Feature activation-4.80e-01

Pos logit contributions
 anonymously+4.898
orr+4.516
oux+4.294
church+4.255
 disse+4.223
Neg logit contributions
advertising-6.373
���-5.895
peria-5.749
 Dresden-5.611
education-5.594
Token election
Token ID3071
Feature activation+1.60e-01

Pos logit contributions
 anonymously+2.959
orr+2.789
oux+2.671
church+2.627
 disse+2.609
Neg logit contributions
advertising-3.635
���-3.356
 Dresden-3.301
peria-3.285
education-3.225
Token and
Token ID290
Feature activation-3.69e-01

Pos logit contributions
advertising+1.112
 Dresden+1.043
���+1.025
ary+1.014
education+1.012
Neg logit contributions
 anonymously-1.056
orr-1.023
oux-0.994
church-0.959
 disse-0.958
Token late
Token ID2739
Feature activation-2.94e-01

Pos logit contributions
 anonymously+0.485
orr+0.470
oux+0.456
church+0.445
 disse+0.445
Neg logit contributions
advertising-0.342
 Dresden-0.317
���-0.310
education-0.302
defense-0.302
Token 2011
Token ID2813
Feature activation+1.33e-01

Pos logit contributions
 anonymously+2.423
orr+2.337
oux+2.263
 disse+2.218
church+2.209
Neg logit contributions
advertising-1.587
���-1.466
 Dresden-1.436
education-1.401
defense-1.392
Token.
Token ID13
Feature activation+3.30e-01

Pos logit contributions
advertising+0.946
 Dresden+0.882
���+0.861
ary+0.858
defense+0.855
Neg logit contributions
 anonymously-0.864
orr-0.831
oux-0.802
 disse-0.779
church-0.774
Token It
Token ID632
Feature activation+1.93e-01

Pos logit contributions
advertising+2.120
 Dresden+2.010
ary+1.902
defense+1.881
education+1.860
Neg logit contributions
 anonymously-2.376
orr-2.275
oux-2.242
 disse-2.212
church-2.162
Token took
Token ID1718
Feature activation+1.79e-01

Pos logit contributions
advertising+1.275
 Dresden+1.181
ary+1.167
���+1.142
defense+1.140
Neg logit contributions
 anonymously-1.336
orr-1.302
oux-1.268
church-1.234
 disse-1.226
Token almost
Token ID2048
Feature activation+1.82e+00

Pos logit contributions
advertising+1.150
 Dresden+1.098
ary+1.059
defense+1.050
education+1.041
Neg logit contributions
 anonymously-1.253
orr-1.215
oux-1.182
church-1.148
 disse-1.140
Token 2
Token ID362
Feature activation+1.17e+00

Pos logit contributions
 Dresden+7.687
advertising+7.338
ary+7.110
 Morales+6.742
defense+6.616
Neg logit contributions
orr-13.804
oux-13.289
church-13.150
nect-13.066
ococ-13.028
Token years
Token ID812
Feature activation+2.80e-01

Pos logit contributions
advertising+6.762
 Dresden+6.284
ary+6.139
defense+5.978
education+5.869
Neg logit contributions
 anonymously-8.141
orr-7.898
oux-7.882
church-7.683
 disse-7.648
Token for
Token ID329
Feature activation+2.49e-01

Pos logit contributions
advertising+1.943
 Dresden+1.824
ary+1.788
defense+1.771
education+1.763
Neg logit contributions
 anonymously-1.840
orr-1.779
oux-1.726
church-1.673
 disse-1.670
Token Bitcoin
Token ID6185
Feature activation-4.06e-01

Pos logit contributions
advertising+1.701
 Dresden+1.621
ary+1.562
defense+1.556
���+1.555
Neg logit contributions
 anonymously-1.645
orr-1.596
oux-1.553
church-1.502
 disse-1.490
Token prices
Token ID4536
Feature activation+4.80e-03

Pos logit contributions
 anonymously+2.754
orr+2.606
oux+2.493
 disse+2.458
church+2.427
Neg logit contributions
advertising-2.791
���-2.656
 Dresden-2.576
education-2.530
peria-2.528
Token promote
Token ID7719
Feature activation-7.66e-03

Pos logit contributions
 anonymously+1.739
orr+1.659
oux+1.594
 disse+1.550
church+1.548
Neg logit contributions
advertising-1.935
���-1.811
 Dresden-1.806
education-1.756
defense-1.753
Token public
Token ID1171
Feature activation+4.09e-01

Pos logit contributions
 anonymously+0.052
orr+0.050
oux+0.048
 disse+0.047
church+0.047
Neg logit contributions
advertising-0.052
 Dresden-0.049
���-0.048
education-0.048
defense-0.047
Token health
Token ID1535
Feature activation-6.65e-02

Pos logit contributions
advertising+2.908
 Dresden+2.730
education+2.711
ary+2.664
defense+2.651
Neg logit contributions
 anonymously-2.598
orr-2.520
oux-2.448
 disse-2.314
church-2.309
Token,"
Token ID553
Feature activation-4.82e-02

Pos logit contributions
 anonymously+0.447
orr+0.430
oux+0.415
 disse+0.403
church+0.403
Neg logit contributions
advertising-0.459
 Dresden-0.428
���-0.424
education-0.416
defense-0.415
Token said
Token ID531
Feature activation-7.49e-02

Pos logit contributions
 anonymously+0.321
orr+0.310
oux+0.298
 disse+0.290
church+0.289
Neg logit contributions
advertising-0.335
 Dresden-0.315
���-0.309
defense-0.305
education-0.305
Token Diane
Token ID26542
Feature activation+1.82e+00

Pos logit contributions
 anonymously+0.511
orr+0.491
oux+0.474
 disse+0.461
church+0.460
Neg logit contributions
advertising-0.510
 Dresden-0.478
���-0.471
education-0.462
defense-0.461
Token Brown
Token ID4373
Feature activation+1.30e-01

Pos logit contributions
 Dresden+10.516
 Morales+10.109
advertising+9.935
ary+9.822
lance+8.820
Neg logit contributions
 anonymously-11.903
 disse-10.983
orr-10.890
oux-10.674
ococ-10.430
Token,
Token ID11
Feature activation-8.84e-03

Pos logit contributions
advertising+0.979
 Dresden+0.923
���+0.911
ary+0.898
education+0.896
Neg logit contributions
 anonymously-0.787
orr-0.748
oux-0.715
 disse-0.706
church-0.691
Token executive
Token ID4640
Feature activation-2.01e-01

Pos logit contributions
 anonymously+0.055
orr+0.053
oux+0.051
church+0.050
 disse+0.050
Neg logit contributions
advertising-0.065
 Dresden-0.061
���-0.060
education-0.059
defense-0.059
Token director
Token ID3437
Feature activation-3.86e-01

Pos logit contributions
 anonymously+1.356
orr+1.293
oux+1.251
 disse+1.214
church+1.212
Neg logit contributions
advertising-1.396
���-1.302
 Dresden-1.299
education-1.259
defense-1.256
Token of
Token ID286
Feature activation-2.63e-01

Pos logit contributions
 anonymously+2.573
orr+2.441
oux+2.379
church+2.292
 disse+2.257
Neg logit contributions
advertising-2.697
���-2.497
 Dresden-2.476
peria-2.415
education-2.393
Tokenun
Token ID403
Feature activation-4.33e-01

Pos logit contributions
 anonymously+4.421
orr+4.283
oux+4.158
church+4.014
ucci+3.878
Neg logit contributions
advertising-3.964
���-3.848
 Dresden-3.749
peria-3.582
defense-3.560
Tokenzel
Token ID17396
Feature activation+2.87e-02

Pos logit contributions
 anonymously+2.379
orr+2.265
oux+2.185
church+2.082
 disse+2.067
Neg logit contributions
advertising-3.537
���-3.365
 Dresden-3.313
peria-3.240
education-3.220
Token?"
Token ID1701
Feature activation+7.11e-01

Pos logit contributions
advertising+0.217
 Dresden+0.205
���+0.201
ary+0.199
defense+0.198
Neg logit contributions
 anonymously-0.174
orr-0.166
oux-0.159
 disse-0.155
church-0.154
Token I
Token ID314
Feature activation+5.97e-01

Pos logit contributions
advertising+4.331
 Dresden+4.116
ary+3.980
defense+3.964
education+3.896
Neg logit contributions
 anonymously-5.034
orr-4.901
oux-4.724
church-4.644
 disse-4.624
Token finally
Token ID3443
Feature activation+1.02e+00

Pos logit contributions
advertising+3.803
 Dresden+3.692
ary+3.631
defense+3.551
education+3.471
Neg logit contributions
orr-3.896
 anonymously-3.873
church-3.802
oux-3.782
 disse-3.643
Token found
Token ID1043
Feature activation+1.82e+00

Pos logit contributions
ary+6.258
 Dresden+6.171
advertising+6.166
defense+6.025
education+5.885
Neg logit contributions
church-6.335
orr-6.331
 anonymously-6.192
oux-6.114
 disse-5.947
Token my
Token ID616
Feature activation+1.38e+00

Pos logit contributions
 Dresden+8.905
advertising+8.494
ary+8.394
defense+8.079
 Morales+7.961
Neg logit contributions
orr-11.975
church-11.899
 anonymously-11.667
oux-11.514
 disse-11.484
Token voice
Token ID3809
Feature activation+6.60e-01

Pos logit contributions
advertising+7.720
 Dresden+7.691
ary+7.500
defense+7.376
education+7.046
Neg logit contributions
orr-9.119
 anonymously-9.026
church-8.598
oux-8.594
kef-8.435
Token and
Token ID290
Feature activation+7.19e-01

Pos logit contributions
advertising+4.145
 Dresden+3.989
ary+3.925
defense+3.891
education+3.839
Neg logit contributions
 anonymously-4.399
orr-4.398
oux-4.227
church-4.182
 disse-4.114
Token gave
Token ID2921
Feature activation+6.16e-01

Pos logit contributions
 Dresden+4.712
advertising+4.711
ary+4.550
defense+4.461
���+4.401
Neg logit contributions
orr-4.481
 anonymously-4.447
oux-4.269
church-4.250
 disse-4.100
Token her
Token ID607
Feature activation+5.91e-01

Pos logit contributions
advertising+3.806
 Dresden+3.662
ary+3.511
defense+3.490
���+3.488
Neg logit contributions
 anonymously-4.337
orr-4.240
oux-4.091
church-4.018
 disse-3.998
Token53
Token ID4310
Feature activation+4.52e-01

Pos logit contributions
advertising+2.465
 Dresden+2.263
ary+2.167
defense+2.156
education+2.135
Neg logit contributions
 anonymously-2.922
orr-2.786
 disse-2.748
oux-2.722
church-2.630
Token p
Token ID279
Feature activation+3.85e-02

Pos logit contributions
advertising+3.433
 Dresden+3.252
ary+3.174
defense+3.149
peria+3.145
Neg logit contributions
 anonymously-2.611
orr-2.482
oux-2.409
church-2.340
 disse-2.336
Token.
Token ID13
Feature activation-3.68e-02

Pos logit contributions
advertising+0.269
 Dresden+0.250
ary+0.243
education+0.243
defense+0.243
Neg logit contributions
 anonymously-0.258
orr-0.247
oux-0.238
 disse-0.235
church-0.231
Tokenm
Token ID76
Feature activation+1.15e+00

Pos logit contributions
 anonymously+0.242
orr+0.230
oux+0.222
 disse+0.219
church+0.215
Neg logit contributions
advertising-0.263
 Dresden-0.245
education-0.238
defense-0.238
ary-0.237
Token.
Token ID13
Feature activation+1.05e+00

Pos logit contributions
advertising+6.314
 Dresden+5.832
defense+5.814
ary+5.796
ebted+5.670
Neg logit contributions
 anonymously-8.187
orr-7.967
 disse-7.794
church-7.631
oux-7.613
Token PT
Token ID19310
Feature activation+1.80e+00

Pos logit contributions
advertising+7.875
 Dresden+7.807
defense+7.578
ary+7.518
 Morales+7.358
Neg logit contributions
 anonymously-5.105
orr-4.954
 disse-4.765
oux-4.738
church-4.638
Token with
Token ID351
Feature activation+7.84e-01

Pos logit contributions
advertising+10.562
ary+10.249
 Dresden+10.244
defense+10.086
 Morales+9.790
Neg logit contributions
orr-9.801
 anonymously-9.796
 disse-9.590
oux-9.234
kef-9.234
Token additional
Token ID3224
Feature activation+6.85e-01

Pos logit contributions
advertising+4.943
 Dresden+4.713
ary+4.539
���+4.469
defense+4.448
Neg logit contributions
 anonymously-5.332
orr-5.287
oux-5.034
church-4.946
 disse-4.898
Token details
Token ID3307
Feature activation+2.43e-01

Pos logit contributions
advertising+4.762
 Dresden+4.509
ary+4.390
defense+4.311
education+4.283
Neg logit contributions
 anonymously-4.250
orr-4.203
oux-4.074
church-3.947
 disse-3.860
Token and
Token ID290
Feature activation+2.47e-01

Pos logit contributions
advertising+1.717
 Dresden+1.608
ary+1.573
���+1.565
education+1.555
Neg logit contributions
 anonymously-1.559
orr-1.509
oux-1.465
church-1.424
 disse-1.417
Token analysis
Token ID3781
Feature activation+4.14e-01

Pos logit contributions
advertising+1.675
 Dresden+1.575
���+1.532
ary+1.525
defense+1.522
Neg logit contributions
 anonymously-1.650
orr-1.614
oux-1.561
church-1.514
 disse-1.492
Token returned
Token ID4504
Feature activation+2.64e-01

Pos logit contributions
 anonymously+0.292
orr+0.281
oux+0.271
church+0.263
 disse+0.262
Neg logit contributions
advertising-0.308
 Dresden-0.286
���-0.281
education-0.279
defense-0.278
Token to
Token ID284
Feature activation+1.86e-01

Pos logit contributions
advertising+1.717
 Dresden+1.624
ary+1.558
education+1.554
defense+1.550
Neg logit contributions
 anonymously-1.838
orr-1.798
oux-1.730
 disse-1.692
church-1.688
Token Al
Token ID978
Feature activation+2.39e-01

Pos logit contributions
advertising+1.219
 Dresden+1.141
���+1.112
education+1.101
defense+1.093
Neg logit contributions
 anonymously-1.312
orr-1.261
oux-1.225
church-1.191
 disse-1.188
Tokencat
Token ID9246
Feature activation-3.76e-01

Pos logit contributions
advertising+1.412
 Dresden+1.280
ary+1.254
defense+1.247
education+1.246
Neg logit contributions
 anonymously-1.886
orr-1.773
 disse-1.737
oux-1.730
church-1.671
Tokenraz
Token ID3247
Feature activation+2.32e-01

Pos logit contributions
 anonymously+3.157
orr+3.107
oux+3.040
church+2.938
 disse+2.863
Neg logit contributions
���-1.912
advertising-1.900
 Dresden-1.736
education-1.654
defense-1.645
Token nearly
Token ID3016
Feature activation+1.80e+00

Pos logit contributions
advertising+1.481
 Dresden+1.382
ary+1.343
education+1.336
defense+1.332
Neg logit contributions
 anonymously-1.644
orr-1.605
oux-1.543
church-1.515
 disse-1.513
Token 50
Token ID2026
Feature activation+9.50e-01

Pos logit contributions
 Dresden+7.973
advertising+7.577
 Morales+7.450
ary+7.319
education+6.881
Neg logit contributions
orr-13.421
oux-13.076
ococ-12.628
church-12.403
nect-12.241
Token years
Token ID812
Feature activation+4.24e-01

Pos logit contributions
advertising+4.958
 Dresden+4.382
defense+4.377
ary+4.278
education+4.228
Neg logit contributions
 anonymously-7.550
orr-7.357
oux-7.146
church-6.986
 disse-6.965
Token after
Token ID706
Feature activation-6.67e-01

Pos logit contributions
advertising+2.461
 Dresden+2.283
ary+2.256
���+2.213
defense+2.201
Neg logit contributions
 anonymously-3.230
orr-3.133
oux-3.070
church-2.994
 disse-2.973
Token the
Token ID262
Feature activation-3.12e-01

Pos logit contributions
 anonymously+4.137
orr+3.945
oux+3.786
 disse+3.723
church+3.674
Neg logit contributions
advertising-4.823
���-4.571
peria-4.432
 Dresden-4.428
ebted-4.408
Token prison
Token ID3770
Feature activation-5.76e-01

Pos logit contributions
 anonymously+2.112
orr+2.025
oux+1.951
 disse+1.900
church+1.896
Neg logit contributions
advertising-2.147
���-1.993
 Dresden-1.993
education-1.938
defense-1.936
Token much
Token ID881
Feature activation-3.11e-01

Pos logit contributions
advertising+1.915
 Dresden+1.783
ary+1.718
���+1.717
education+1.716
Neg logit contributions
 anonymously-1.900
orr-1.843
oux-1.787
church-1.742
 disse-1.729
Token as
Token ID355
Feature activation+9.03e-02

Pos logit contributions
 anonymously+1.336
orr+1.256
oux+1.175
 disse+1.129
church+1.114
Neg logit contributions
advertising-2.909
���-2.761
 Dresden-2.758
education-2.701
peria-2.697
Token he
Token ID339
Feature activation+1.58e-01

Pos logit contributions
advertising+0.630
 Dresden+0.590
���+0.583
education+0.571
defense+0.570
Neg logit contributions
 anonymously-0.601
orr-0.578
oux-0.556
church-0.542
 disse-0.541
Token
Token ID564
Feature activation+5.68e-01

Pos logit contributions
advertising+1.089
 Dresden+1.013
���+0.992
education+0.983
ary+0.983
Neg logit contributions
 anonymously-1.057
orr-1.016
oux-0.983
church-0.966
 disse-0.954
Token
Token ID250
Feature activation-8.90e-02

Pos logit contributions
advertising+2.390
 Dresden+2.128
education+2.025
defense+2.022
ary+1.949
Neg logit contributions
 anonymously-5.287
 disse-5.048
orr-5.034
oux-5.009
church-4.883
Tokenw
Token ID86
Feature activation+1.78e+00

Pos logit contributions
 anonymously+0.678
orr+0.653
oux+0.632
 disse+0.617
church+0.616
Neg logit contributions
advertising-0.539
 Dresden-0.496
���-0.491
education-0.481
defense-0.479
Tokenants
Token ID1187
Feature activation+5.08e-01

Pos logit contributions
ary+10.890
advertising+10.146
education+9.879
lance+9.467
amily+9.286
Neg logit contributions
 anonymously-11.583
church-9.953
 disse-9.933
oux-9.707
orr-9.685
Token to
Token ID284
Feature activation-3.87e-02

Pos logit contributions
advertising+3.175
 Dresden+2.958
defense+2.854
ary+2.851
education+2.838
Neg logit contributions
 anonymously-3.646
orr-3.553
oux-3.429
church-3.402
 disse-3.334
Token be
Token ID307
Feature activation+1.57e-02

Pos logit contributions
 anonymously+0.249
orr+0.240
oux+0.232
church+0.227
 disse+0.223
Neg logit contributions
advertising-0.276
 Dresden-0.260
���-0.254
ary-0.252
defense-0.252
Token able
Token ID1498
Feature activation+3.78e-01

Pos logit contributions
advertising+0.118
 Dresden+0.110
���+0.108
ary+0.107
education+0.107
Neg logit contributions
 anonymously-0.097
orr-0.093
oux-0.090
church-0.087
 disse-0.086
Token to
Token ID284
Feature activation+8.59e-02

Pos logit contributions
advertising+2.311
 Dresden+2.139
���+2.125
ary+2.080
education+2.067
Neg logit contributions
 anonymously-2.795
orr-2.712
oux-2.625
church-2.568
 disse-2.543
Token which
Token ID543
Feature activation+5.76e-01

Pos logit contributions
 anonymously+0.227
orr+0.219
oux+0.210
church+0.205
 disse+0.204
Neg logit contributions
advertising-0.255
 Dresden-0.240
���-0.237
education-0.232
ary-0.231
Token I
Token ID314
Feature activation-2.13e-01

Pos logit contributions
advertising+3.541
 Dresden+3.468
ary+3.309
���+3.238
education+3.199
Neg logit contributions
 anonymously-3.988
orr-3.912
oux-3.802
church-3.730
 disse-3.645
Token have
Token ID423
Feature activation-2.88e-01

Pos logit contributions
 anonymously+1.357
orr+1.302
oux+1.253
church+1.218
 disse+1.215
Neg logit contributions
advertising-1.548
 Dresden-1.453
���-1.441
education-1.410
defense-1.407
Token kept
Token ID4030
Feature activation+4.09e-01

Pos logit contributions
 anonymously+1.778
orr+1.687
oux+1.617
 disse+1.571
church+1.566
Neg logit contributions
advertising-2.165
 Dresden-2.021
���-2.020
education-1.975
defense-1.974
Token for
Token ID329
Feature activation+1.07e+00

Pos logit contributions
advertising+2.699
 Dresden+2.639
ary+2.559
���+2.483
education+2.473
Neg logit contributions
orr-2.635
 anonymously-2.609
oux-2.568
church-2.471
 disse-2.459
Token almost
Token ID2048
Feature activation+1.78e+00

Pos logit contributions
 Dresden+6.430
advertising+6.208
ary+6.051
education+5.903
���+5.795
Neg logit contributions
orr-6.841
 anonymously-6.837
oux-6.834
church-6.674
nect-6.612
Token 30
Token ID1542
Feature activation+2.70e-01

Pos logit contributions
 Dresden+7.568
advertising+6.987
ary+6.895
education+6.467
peria+6.313
Neg logit contributions
orr-13.407
oux-13.128
church-13.054
nect-12.935
ucci-12.897
Token years
Token ID812
Feature activation+3.35e-01

Pos logit contributions
advertising+1.728
 Dresden+1.613
���+1.573
defense+1.549
ary+1.549
Neg logit contributions
 anonymously-1.929
orr-1.857
oux-1.798
 disse-1.757
church-1.757
Token.
Token ID13
Feature activation+2.88e-01

Pos logit contributions
advertising+2.325
 Dresden+2.197
ary+2.147
���+2.110
education+2.107
Neg logit contributions
 anonymously-2.143
orr-2.096
oux-2.050
church-1.985
 disse-1.971
Token
Token ID198
Feature activation+1.12e-01

Pos logit contributions
advertising+1.882
 Dresden+1.792
ary+1.697
education+1.681
���+1.667
Neg logit contributions
 anonymously-2.038
orr-1.951
oux-1.893
 disse-1.882
church-1.861
Token
Token ID198
Feature activation+2.13e-01

Pos logit contributions
advertising+0.781
 Dresden+0.722
ary+0.709
education+0.700
defense+0.697
Neg logit contributions
 anonymously-0.761
orr-0.724
 disse-0.705
oux-0.704
church-0.682
Token MORE
Token ID12011
Feature activation-5.00e-01

Pos logit contributions
 anonymously+2.264
orr+2.182
oux+2.113
 disse+2.065
church+2.056
Neg logit contributions
advertising-2.122
 Dresden-1.978
���-1.956
education-1.915
defense-1.910
Token and
Token ID290
Feature activation+2.08e-01

Pos logit contributions
 anonymously+2.923
orr+2.760
oux+2.660
church+2.535
 disse+2.496
Neg logit contributions
advertising-3.902
���-3.755
 Dresden-3.616
peria-3.578
defense-3.555
Token Hillary
Token ID5073
Feature activation+8.46e-01

Pos logit contributions
advertising+1.401
 Dresden+1.326
ary+1.281
���+1.277
education+1.276
Neg logit contributions
 anonymously-1.398
orr-1.369
oux-1.321
church-1.278
 disse-1.277
Token Clinton
Token ID2605
Feature activation+9.06e-01

Pos logit contributions
advertising+4.323
 Dresden+4.191
ary+4.017
 Morales+3.777
education+3.769
Neg logit contributions
 anonymously-6.689
orr-6.591
 disse-6.523
oux-6.473
church-6.322
Token Hillary
Token ID5073
Feature activation-2.54e-01

Pos logit contributions
advertising+4.632
 Dresden+4.449
ary+4.308
education+4.165
peria+4.039
Neg logit contributions
 anonymously-6.994
orr-6.923
church-6.875
oux-6.842
 disse-6.790
Token Diane
Token ID26542
Feature activation+1.78e+00

Pos logit contributions
 anonymously+1.784
orr+1.684
oux+1.632
 disse+1.582
church+1.581
Neg logit contributions
advertising-1.698
���-1.617
 Dresden-1.549
peria-1.531
education-1.525
Token Rodham
Token ID41848
Feature activation+3.97e-01

Pos logit contributions
 Dresden+3.451
advertising+3.305
lance+2.621
ary+2.620
 Morales+2.410
Neg logit contributions
orr-19.423
 anonymously-19.138
oux-18.802
 disse-18.690
church-18.246
Token Clinton
Token ID2605
Feature activation-7.34e-02

Pos logit contributions
advertising+2.447
 Dresden+2.283
education+2.183
ary+2.177
defense+2.165
Neg logit contributions
 anonymously-2.962
orr-2.870
oux-2.778
 disse-2.745
church-2.716
TokenSanders
Token ID26747
Feature activation+1.13e-02

Pos logit contributions
 anonymously+0.472
orr+0.449
oux+0.430
church+0.418
 disse+0.413
Neg logit contributions
advertising-0.524
���-0.500
 Dresden-0.496
peria-0.478
education-0.478
Token:
Token ID25
Feature activation-2.39e-02

Pos logit contributions
advertising+0.075
 Dresden+0.070
���+0.069
education+0.068
defense+0.068
Neg logit contributions
 anonymously-0.079
orr-0.076
oux-0.074
 disse-0.071
church-0.071
Token '
Token ID705
Feature activation+3.33e-01

Pos logit contributions
 anonymously+0.187
orr+0.181
oux+0.175
 disse+0.171
church+0.171
Neg logit contributions
advertising-0.139
 Dresden-0.128
���-0.127
education-0.123
defense-0.123
Token war
Token ID1175
Feature activation+2.97e-01

Pos logit contributions
advertising+2.223
 Dresden+2.133
ary+2.083
education+2.067
defense+2.064
Neg logit contributions
 anonymously-2.276
orr-2.197
oux-2.152
church-2.068
 disse-2.027
Token
Token ID447
Feature activation+7.41e-03

Pos logit contributions
advertising+1.958
 Dresden+1.827
ary+1.781
���+1.773
defense+1.772
Neg logit contributions
 anonymously-2.062
orr-1.980
oux-1.929
 disse-1.881
church-1.870
Token
Token ID251
Feature activation+1.13e-01

Pos logit contributions
advertising+0.040
 Dresden+0.037
���+0.036
education+0.036
defense+0.035
Neg logit contributions
 anonymously-0.061
orr-0.058
oux-0.057
 disse-0.056
church-0.055
Token and
Token ID290
Feature activation+5.93e-01

Pos logit contributions
advertising+0.819
 Dresden+0.767
���+0.752
education+0.747
defense+0.746
Neg logit contributions
 anonymously-0.725
orr-0.693
oux-0.669
 disse-0.652
church-0.646
Token the
Token ID262
Feature activation+1.92e-01

Pos logit contributions
advertising+3.836
 Dresden+3.694
ary+3.537
education+3.522
defense+3.511
Neg logit contributions
 anonymously-3.947
orr-3.903
oux-3.804
church-3.733
 disse-3.611
Token contr
Token ID3445
Feature activation+1.78e+00

Pos logit contributions
advertising+1.328
 Dresden+1.260
���+1.227
education+1.212
defense+1.211
Neg logit contributions
 anonymously-1.271
orr-1.224
oux-1.183
 disse-1.145
church-1.144
Tokenaband
Token ID49248
Feature activation-1.35e-01

Pos logit contributions
advertising+8.589
ary+8.575
education+7.482
amily+7.448
defense+7.315
Neg logit contributions
 anonymously-13.401
 disse-12.441
oux-12.437
church-12.330
orr-12.144
Token trade
Token ID3292
Feature activation-3.68e-01

Pos logit contributions
 anonymously+0.895
orr+0.858
oux+0.826
 disse+0.803
church+0.802
Neg logit contributions
advertising-0.942
 Dresden-0.879
���-0.875
education-0.853
defense-0.852
Token.
Token ID13
Feature activation+1.84e-01

Pos logit contributions
 anonymously+2.361
orr+2.246
oux+2.150
 disse+2.086
church+2.081
Neg logit contributions
advertising-2.659
���-2.515
 Dresden-2.476
peria-2.422
education-2.413
Token
Token ID198
Feature activation-1.30e-02

Pos logit contributions
advertising+1.251
 Dresden+1.172
defense+1.124
���+1.122
ary+1.122
Neg logit contributions
 anonymously-1.269
orr-1.214
oux-1.177
 disse-1.157
church-1.145
Token
Token ID198
Feature activation+2.26e-01

Pos logit contributions
 anonymously+0.085
orr+0.080
oux+0.078
 disse+0.077
church+0.075
Neg logit contributions
advertising-0.094
 Dresden-0.087
education-0.085
ary-0.085
defense-0.085
Tokenised
Token ID1417
Feature activation-2.26e-01

Pos logit contributions
 anonymously+3.239
oux+2.953
orr+2.939
church+2.663
 disse+2.589
Neg logit contributions
advertising-9.148
���-9.083
 Dresden-8.933
defense-8.593
peria-8.469
Token on
Token ID319
Feature activation+4.31e-01

Pos logit contributions
 anonymously+1.308
orr+1.219
oux+1.180
 disse+1.142
church+1.132
Neg logit contributions
advertising-1.786
 Dresden-1.689
���-1.685
education-1.637
defense-1.637
Token a
Token ID257
Feature activation+1.70e-01

Pos logit contributions
advertising+2.863
 Dresden+2.678
ary+2.599
education+2.594
���+2.584
Neg logit contributions
 anonymously-2.914
orr-2.854
oux-2.750
church-2.685
 disse-2.636
Token vast
Token ID5909
Feature activation-1.03e-01

Pos logit contributions
advertising+1.183
 Dresden+1.115
���+1.099
ary+1.081
education+1.076
Neg logit contributions
 anonymously-1.119
orr-1.078
oux-1.035
church-1.008
 disse-1.002
Token m
Token ID285
Feature activation-3.63e-02

Pos logit contributions
 anonymously+0.675
orr+0.647
oux+0.623
 disse+0.604
church+0.604
Neg logit contributions
advertising-0.731
 Dresden-0.683
���-0.679
education-0.663
defense-0.662
Tokeniscal
Token ID7860
Feature activation-2.08e+00

Pos logit contributions
 anonymously+0.259
orr+0.249
oux+0.241
 disse+0.235
church+0.235
Neg logit contributions
advertising-0.236
 Dresden-0.219
���-0.217
education-0.212
defense-0.212
Tokenculation
Token ID14902
Feature activation-3.22e-01

Pos logit contributions
orr+8.876
 anonymously+8.595
oux+7.861
church+7.859
ucci+7.712
Neg logit contributions
 Dresden-15.569
���-15.522
advertising-15.088
peria-14.411
defense-14.380
Token about
Token ID546
Feature activation+7.78e-02

Pos logit contributions
 anonymously+2.166
orr+2.070
oux+1.987
 disse+1.938
church+1.931
Neg logit contributions
advertising-2.221
���-2.092
 Dresden-2.067
education-2.014
defense-2.013
Token the
Token ID262
Feature activation+3.80e-01

Pos logit contributions
advertising+0.526
 Dresden+0.493
education+0.479
���+0.478
ary+0.476
Neg logit contributions
 anonymously-0.528
orr-0.515
oux-0.495
 disse-0.479
church-0.478
Token origins
Token ID15587
Feature activation+3.95e-01

Pos logit contributions
advertising+2.458
 Dresden+2.329
ary+2.268
education+2.249
defense+2.219
Neg logit contributions
 anonymously-2.624
orr-2.551
oux-2.439
church-2.371
 disse-2.363
Token of
Token ID286
Feature activation+1.76e-01

Pos logit contributions
advertising+2.567
 Dresden+2.377
���+2.357
ary+2.314
education+2.314
Neg logit contributions
 anonymously-2.807
orr-2.700
oux-2.605
 disse-2.543
church-2.533
Token where
Token ID810
Feature activation+5.39e-01

Pos logit contributions
 anonymously+0.119
orr+0.115
oux+0.111
church+0.107
 disse+0.107
Neg logit contributions
advertising-0.123
 Dresden-0.116
���-0.114
education-0.112
defense-0.112
Token do
Token ID466
Feature activation+1.51e-01

Pos logit contributions
advertising+3.609
 Dresden+3.402
education+3.297
ary+3.278
���+3.248
Neg logit contributions
 anonymously-3.596
orr-3.498
oux-3.367
 disse-3.281
church-3.265
Token we
Token ID356
Feature activation-8.75e-01

Pos logit contributions
advertising+1.034
 Dresden+0.977
���+0.956
education+0.941
ary+0.939
Neg logit contributions
 anonymously-1.009
orr-0.975
oux-0.939
church-0.908
 disse-0.907
Token draw
Token ID3197
Feature activation-3.70e-01

Pos logit contributions
 anonymously+5.018
orr+4.648
oux+4.431
 disse+4.322
church+4.193
Neg logit contributions
advertising-6.988
���-6.682
 Dresden-6.406
peria-6.359
ebted-6.307
Token the
Token ID262
Feature activation-6.75e-01

Pos logit contributions
 anonymously+2.420
orr+2.288
oux+2.215
 disse+2.161
church+2.139
Neg logit contributions
advertising-2.641
���-2.465
 Dresden-2.434
education-2.384
peria-2.379
Token line
Token ID1627
Feature activation-2.06e+00

Pos logit contributions
 anonymously+4.144
orr+3.893
oux+3.782
 disse+3.720
church+3.623
Neg logit contributions
advertising-5.029
���-4.731
 Dresden-4.592
peria-4.555
education-4.539
Token?
Token ID30
Feature activation-3.86e-01

Pos logit contributions
 anonymously+10.218
orr+8.787
oux+8.786
 disse+8.509
church+7.988
Neg logit contributions
���-14.829
advertising-14.237
peria-13.553
education-13.428
amily-13.330
Token So
Token ID1406
Feature activation-9.31e-02

Pos logit contributions
 anonymously+2.543
orr+2.435
oux+2.344
church+2.277
 disse+2.262
Neg logit contributions
advertising-2.668
���-2.546
 Dresden-2.477
peria-2.438
defense-2.432
Token they
Token ID484
Feature activation+6.18e-02

Pos logit contributions
 anonymously+0.605
orr+0.581
oux+0.560
 disse+0.543
church+0.543
Neg logit contributions
advertising-0.665
 Dresden-0.622
���-0.614
education-0.603
ary-0.601
Token say
Token ID910
Feature activation-2.09e-01

Pos logit contributions
advertising+0.435
 Dresden+0.411
���+0.402
ary+0.400
education+0.398
Neg logit contributions
 anonymously-0.396
orr-0.386
oux-0.374
church-0.363
 disse-0.357
Token,
Token ID11
Feature activation+2.74e-02

Pos logit contributions
 anonymously+1.331
orr+1.272
oux+1.222
 disse+1.187
church+1.186
Neg logit contributions
advertising-1.525
 Dresden-1.425
���-1.423
education-1.387
defense-1.385
Token/
Token ID14
Feature activation+1.43e-01

Pos logit contributions
advertising+2.260
 Dresden+2.080
ary+2.049
education+2.023
defense+2.023
Neg logit contributions
 anonymously-2.447
orr-2.387
oux-2.286
 disse-2.239
church-2.198
TokenOffice
Token ID27743
Feature activation+2.90e-01

Pos logit contributions
advertising+0.951
 Dresden+0.868
education+0.854
���+0.854
defense+0.849
Neg logit contributions
 anonymously-1.005
orr-0.969
oux-0.932
 disse-0.913
church-0.899
Token)
Token ID8
Feature activation+4.43e-01

Pos logit contributions
advertising+1.907
 Dresden+1.770
���+1.739
ary+1.729
education+1.728
Neg logit contributions
 anonymously-2.010
orr-1.960
oux-1.887
 disse-1.833
church-1.825
Token and
Token ID290
Feature activation+1.60e-01

Pos logit contributions
advertising+3.006
 Dresden+2.814
education+2.740
���+2.723
ary+2.709
Neg logit contributions
 anonymously-2.909
orr-2.882
oux-2.754
church-2.667
 disse-2.650
Token enable
Token ID7139
Feature activation+2.00e-01

Pos logit contributions
advertising+1.111
 Dresden+1.039
���+1.018
education+1.007
ary+1.002
Neg logit contributions
 anonymously-1.060
orr-1.038
oux-0.991
church-0.964
 disse-0.962
Token Broad
Token ID9765
Feature activation-2.05e+00

Pos logit contributions
advertising+1.401
 Dresden+1.308
���+1.284
ary+1.268
education+1.265
Neg logit contributions
 anonymously-1.318
orr-1.282
oux-1.225
church-1.192
 disse-1.185
Tokenband
Token ID3903
Feature activation-1.03e-01

Pos logit contributions
 anonymously+9.868
church+9.513
orr+9.175
oux+8.591
nect+8.516
Neg logit contributions
���-15.233
 Dresden-14.337
advertising-13.939
defense-13.779
ebted-13.711
Token Connection
Token ID26923
Feature activation+2.97e-01

Pos logit contributions
 anonymously+0.728
orr+0.702
oux+0.677
 disse+0.659
church+0.659
Neg logit contributions
advertising-0.676
 Dresden-0.627
���-0.621
education-0.608
defense-0.607
Token (
Token ID357
Feature activation-1.90e-02

Pos logit contributions
advertising+1.939
 Dresden+1.803
ary+1.769
education+1.745
defense+1.730
Neg logit contributions
 anonymously-2.051
orr-2.023
oux-1.939
church-1.903
 disse-1.886
TokenCo
Token ID7222
Feature activation-1.64e-01

Pos logit contributions
 anonymously+0.135
orr+0.130
oux+0.125
 disse+0.123
church+0.121
Neg logit contributions
advertising-0.126
 Dresden-0.115
���-0.114
education-0.113
defense-0.113
Tokenax
Token ID897
Feature activation+7.63e-01

Pos logit contributions
 anonymously+1.253
orr+1.211
oux+1.171
church+1.142
 disse+1.136
Neg logit contributions
advertising-0.986
���-0.918
 Dresden-0.915
defense-0.880
education-0.874
Token a
Token ID257
Feature activation+1.94e-01

Pos logit contributions
advertising+2.524
 Dresden+2.386
���+2.317
ary+2.270
education+2.254
Neg logit contributions
 anonymously-3.028
orr-2.949
oux-2.849
church-2.799
 disse-2.752
Token nightmare
Token ID17123
Feature activation+7.06e-01

Pos logit contributions
advertising+1.317
 Dresden+1.253
���+1.215
ary+1.195
defense+1.194
Neg logit contributions
 anonymously-1.303
orr-1.261
oux-1.213
church-1.183
 disse-1.168
Token apartment
Token ID7962
Feature activation-2.08e-02

Pos logit contributions
advertising+4.443
 Dresden+4.309
ary+4.195
defense+4.047
education+4.041
Neg logit contributions
 anonymously-4.852
orr-4.753
oux-4.575
church-4.477
 disse-4.381
Token broker
Token ID20426
Feature activation+2.37e-01

Pos logit contributions
 anonymously+0.142
orr+0.137
oux+0.132
 disse+0.128
church+0.128
Neg logit contributions
advertising-0.141
 Dresden-0.132
���-0.131
education-0.128
defense-0.127
Token on
Token ID319
Feature activation+3.23e-02

Pos logit contributions
advertising+1.630
 Dresden+1.542
���+1.498
ary+1.494
education+1.484
Neg logit contributions
 anonymously-1.564
orr-1.513
oux-1.458
church-1.422
 disse-1.406
Token Broad
Token ID9765
Feature activation-2.03e+00

Pos logit contributions
advertising+0.214
 Dresden+0.200
���+0.197
education+0.192
defense+0.192
Neg logit contributions
 anonymously-0.227
orr-0.218
oux-0.211
church-0.205
 disse-0.205
Token City
Token ID2254
Feature activation+5.88e-04

Pos logit contributions
church+9.666
 anonymously+8.033
orr+7.305
oux+6.767
 disse+6.756
Neg logit contributions
���-16.576
advertising-16.180
defense-15.643
 Dresden-15.355
education-15.304
Token and
Token ID290
Feature activation+4.34e-01

Pos logit contributions
advertising+0.004
 Dresden+0.004
���+0.004
education+0.004
defense+0.004
Neg logit contributions
 anonymously-0.004
orr-0.004
oux-0.004
church-0.003
 disse-0.003
Token once
Token ID1752
Feature activation+4.46e-01

Pos logit contributions
advertising+2.840
 Dresden+2.731
���+2.600
ary+2.584
defense+2.562
Neg logit contributions
 anonymously-2.947
orr-2.889
oux-2.796
church-2.729
 disse-2.689
Token did
Token ID750
Feature activation+5.46e-01

Pos logit contributions
advertising+3.055
 Dresden+2.910
ary+2.834
���+2.833
defense+2.770
Neg logit contributions
 anonymously-2.854
orr-2.832
oux-2.717
church-2.679
 disse-2.597
Token cart
Token ID6383
Feature activation-2.78e-01

Pos logit contributions
advertising+3.631
 Dresden+3.455
���+3.315
ary+3.297
education+3.283
Neg logit contributions
 anonymously-3.639
orr-3.532
oux-3.379
church-3.345
 disse-3.286
Token including
Token ID1390
Feature activation-3.61e-01

Pos logit contributions
 anonymously+0.044
orr+0.043
oux+0.041
 disse+0.040
church+0.040
Neg logit contributions
advertising-0.045
 Dresden-0.042
���-0.042
defense-0.041
education-0.041
Token House
Token ID2097
Feature activation-1.12e+00

Pos logit contributions
 anonymously+2.650
orr+2.525
oux+2.443
 disse+2.378
church+2.376
Neg logit contributions
advertising-2.295
 Dresden-2.100
���-2.100
education-2.035
peria-2.031
Token Majority
Token ID22171
Feature activation-9.03e-01

Pos logit contributions
 anonymously+5.781
orr+5.136
oux+4.998
church+4.939
 disse+4.898
Neg logit contributions
advertising-9.165
���-8.652
 Dresden-8.476
peria-8.373
ebted-8.239
Token Whip
Token ID40930
Feature activation-3.31e-01

Pos logit contributions
 anonymously+3.430
orr+3.033
oux+2.857
church+2.683
 disse+2.608
Neg logit contributions
advertising-8.783
���-8.296
 Dresden-8.204
education-8.109
peria-8.098
Token Steve
Token ID6542
Feature activation+1.73e-01

Pos logit contributions
 anonymously+1.568
orr+1.472
oux+1.383
 disse+1.325
church+1.320
Neg logit contributions
advertising-2.960
���-2.806
 Dresden-2.772
defense-2.738
education-2.724
Token Scal
Token ID34529
Feature activation-2.01e+00

Pos logit contributions
advertising+1.111
 Dresden+1.036
���+1.023
education+0.999
defense+0.998
Neg logit contributions
 anonymously-1.249
orr-1.201
oux-1.161
 disse-1.133
church-1.128
Tokenise
Token ID786
Feature activation-5.31e-01

Pos logit contributions
orr+10.340
oux+10.297
ucci+9.162
church+8.698
 anonymously+8.408
Neg logit contributions
advertising-15.266
 Dresden-13.752
���-13.735
defense-13.658
-13.639
Token (
Token ID357
Feature activation-8.07e-01

Pos logit contributions
 anonymously+3.723
orr+3.563
oux+3.475
 disse+3.292
church+3.287
Neg logit contributions
advertising-3.518
���-3.273
 Dresden-3.210
education-3.150
defense-3.139
TokenR
Token ID49
Feature activation-3.67e-01

Pos logit contributions
 anonymously+4.502
orr+4.409
oux+4.275
church+4.176
 disse+3.976
Neg logit contributions
advertising-6.109
���-6.089
 Dresden-5.873
amily-5.635
education-5.584
Token-
Token ID12
Feature activation-8.99e-01

Pos logit contributions
 anonymously+2.338
orr+2.270
oux+2.261
church+2.107
ucci+2.043
Neg logit contributions
advertising-2.616
���-2.478
 Dresden-2.474
defense-2.398
education-2.338
TokenLa
Token ID14772
Feature activation+1.90e-01

Pos logit contributions
 anonymously+5.368
orr+5.169
oux+5.032
church+4.936
 disse+4.783
Neg logit contributions
advertising-6.350
���-6.294
 Dresden-6.161
amily-5.924
education-5.876
Token that
Token ID326
Feature activation+8.31e-01

Pos logit contributions
advertising+4.583
 Dresden+4.453
ary+4.388
defense+4.183
 Morales+4.117
Neg logit contributions
 anonymously-5.218
orr-5.120
oux-5.112
church-4.908
 disse-4.875
Token today
Token ID1909
Feature activation+6.91e-01

Pos logit contributions
advertising+5.107
 Dresden+4.982
ary+4.876
defense+4.675
 Morales+4.653
Neg logit contributions
oux-5.434
 anonymously-5.421
orr-5.417
church-5.218
nect-5.089
Token's
Token ID338
Feature activation+3.28e-01

Pos logit contributions
advertising+4.164
 Dresden+3.914
ary+3.858
defense+3.743
education+3.696
Neg logit contributions
 anonymously-4.981
orr-4.822
oux-4.779
church-4.590
 disse-4.571
Token crash
Token ID7014
Feature activation+1.00e-01

Pos logit contributions
advertising+2.158
 Dresden+2.057
ary+1.981
���+1.950
defense+1.949
Neg logit contributions
 anonymously-2.227
orr-2.170
oux-2.132
church-2.048
 disse-2.009
Token will
Token ID481
Feature activation+4.97e-01

Pos logit contributions
advertising+0.689
 Dresden+0.646
ary+0.634
���+0.626
defense+0.624
Neg logit contributions
 anonymously-0.669
orr-0.651
oux-0.631
church-0.608
 disse-0.604
Token inaug
Token ID14243
Feature activation-1.99e+00

Pos logit contributions
advertising+3.218
ary+3.154
 Dresden+3.121
defense+3.010
education+2.911
Neg logit contributions
 anonymously-3.193
orr-3.192
oux-3.122
church-3.114
 disse-2.976
Tokenurate
Token ID15537
Feature activation+8.81e-02

Pos logit contributions
orr+6.332
oux+6.156
scribe+5.780
nect+5.432
urated+5.385
Neg logit contributions
���-20.180
advertising-17.901
 Dresden-17.578
 Morales-16.815
defense-16.690
Token a
Token ID257
Feature activation+6.18e-01

Pos logit contributions
advertising+0.457
 Dresden+0.417
���+0.412
education+0.399
defense+0.399
Neg logit contributions
 anonymously-0.745
orr-0.721
oux-0.700
 disse-0.684
church-0.684
Token bear
Token ID6842
Feature activation+5.62e-01

Pos logit contributions
advertising+4.044
 Dresden+3.980
ary+3.880
defense+3.718
���+3.702
Neg logit contributions
 anonymously-4.087
orr-3.964
oux-3.882
church-3.748
 disse-3.680
Token market
Token ID1910
Feature activation+1.90e-01

Pos logit contributions
advertising+2.339
ary+2.186
 Dresden+2.094
defense+2.055
education+1.977
Neg logit contributions
 anonymously-5.162
orr-4.920
oux-4.895
church-4.819
 disse-4.798
Token like
Token ID588
Feature activation-1.56e-01

Pos logit contributions
advertising+1.332
 Dresden+1.254
ary+1.230
defense+1.213
education+1.210
Neg logit contributions
 anonymously-1.236
orr-1.196
oux-1.158
church-1.121
 disse-1.115
Token reached
Token ID4251
Feature activation-2.75e-01

Pos logit contributions
 anonymously+2.667
orr+2.519
oux+2.418
 disse+2.339
church+2.314
Neg logit contributions
advertising-3.529
���-3.325
 Dresden-3.303
defense-3.229
education-3.225
Token at
Token ID379
Feature activation-8.97e-01

Pos logit contributions
 anonymously+1.862
orr+1.765
oux+1.710
church+1.651
 disse+1.649
Neg logit contributions
advertising-1.904
���-1.794
 Dresden-1.778
education-1.724
defense-1.723
Token j
Token ID474
Feature activation-3.83e-01

Pos logit contributions
 anonymously+5.151
orr+4.814
 disse+4.611
oux+4.520
church+4.411
Neg logit contributions
advertising-6.952
���-6.625
 Dresden-6.466
peria-6.385
education-6.356
Tokenon
Token ID261
Feature activation-2.04e-01

Pos logit contributions
 anonymously+2.364
orr+2.341
oux+2.224
church+2.206
 disse+2.036
Neg logit contributions
���-2.866
advertising-2.753
 Dresden-2.738
education-2.525
peria-2.523
Token.
Token ID13
Feature activation-1.57e-01

Pos logit contributions
 anonymously+1.267
orr+1.227
oux+1.164
church+1.160
 disse+1.118
Neg logit contributions
advertising-1.490
 Dresden-1.450
���-1.446
defense-1.360
peria-1.358
Tokenchest
Token ID46713
Feature activation-1.96e+00

Pos logit contributions
 anonymously+1.188
orr+1.154
oux+1.111
church+1.094
 disse+1.076
Neg logit contributions
advertising-0.944
���-0.892
 Dresden-0.885
education-0.840
peria-0.838
Tokeno
Token ID78
Feature activation-8.17e-01

Pos logit contributions
orr+8.514
oux+8.508
 anonymously+7.475
ococ+7.215
church+7.166
Neg logit contributions
���-19.106
 Dresden-16.309
advertising-15.353
 contestant-14.963
lance-14.751
Token@
Token ID31
Feature activation-1.05e+00

Pos logit contributions
 anonymously+4.560
orr+4.292
oux+4.146
church+3.897
 disse+3.806
Neg logit contributions
advertising-6.347
���-6.186
 Dresden-6.076
peria-5.870
defense-5.823
Tokengl
Token ID4743
Feature activation-9.43e-01

Pos logit contributions
 anonymously+5.746
orr+5.660
church+5.393
oux+5.343
 disse+5.067
Neg logit contributions
advertising-7.916
���-7.892
 Dresden-7.487
 Morales-7.247
peria-7.177
Tokenobe
Token ID5910
Feature activation+8.71e-02

Pos logit contributions
orr+4.612
 anonymously+4.528
oux+4.420
church+4.243
ococ+3.775
Neg logit contributions
���-8.356
advertising-8.026
 Dresden-7.816
peria-7.369
eanor-7.314
Token.
Token ID13
Feature activation-1.99e-02

Pos logit contributions
advertising+0.597
 Dresden+0.545
education+0.543
defense+0.539
ary+0.538
Neg logit contributions
 anonymously-0.594
orr-0.564
oux-0.546
 disse-0.546
church-0.531
TokenHA
Token ID7801
Feature activation+5.87e-01

Pos logit contributions
advertising+3.857
 Dresden+3.610
ary+3.543
defense+3.479
peria+3.458
Neg logit contributions
 anonymously-3.644
orr-3.490
oux-3.332
 disse-3.332
church-3.215
TokenW
Token ID54
Feature activation+5.73e-01

Pos logit contributions
advertising+4.077
 Dresden+3.852
ary+3.681
defense+3.665
education+3.606
Neg logit contributions
 anonymously-3.805
orr-3.677
oux-3.516
 disse-3.499
church-3.451
TokenK
Token ID42
Feature activation+1.89e-01

Pos logit contributions
advertising+3.833
 Dresden+3.646
ary+3.557
education+3.426
defense+3.420
Neg logit contributions
 anonymously-3.870
orr-3.694
oux-3.549
 disse-3.527
church-3.487
Token AND
Token ID5357
Feature activation+3.52e-01

Pos logit contributions
advertising+1.324
 Dresden+1.240
���+1.205
ary+1.204
defense+1.197
Neg logit contributions
 anonymously-1.251
orr-1.197
oux-1.156
 disse-1.131
church-1.124
Token AP
Token ID3486
Feature activation-3.17e-01

Pos logit contributions
advertising+2.596
 Dresden+2.454
ary+2.355
���+2.343
defense+2.339
Neg logit contributions
 anonymously-2.182
orr-2.082
oux-2.014
 disse-1.961
church-1.960
TokenAC
Token ID2246
Feature activation-1.96e+00

Pos logit contributions
 anonymously+1.566
orr+1.498
oux+1.415
church+1.357
 disse+1.337
Neg logit contributions
advertising-2.713
���-2.593
 Dresden-2.585
education-2.537
defense-2.517
TokenHE
Token ID13909
Feature activation-5.65e-01

Pos logit contributions
HT+4.438
orr+3.472
 anonymously+3.400
church+3.055
oux+2.710
Neg logit contributions
���-21.970
advertising-18.231
 Dresden-18.055
 contestant-17.838
defense-17.819
Token T
Token ID309
Feature activation+7.87e-02

Pos logit contributions
 anonymously+3.361
orr+3.153
oux+3.053
 disse+2.959
church+2.911
Neg logit contributions
advertising-4.230
���-4.054
education-3.943
 Dresden-3.941
peria-3.904
TokenOC
Token ID4503
Feature activation-4.65e-01

Pos logit contributions
advertising+0.448
 Dresden+0.404
ary+0.392
defense+0.386
education+0.385
Neg logit contributions
 anonymously-0.635
orr-0.606
 disse-0.589
oux-0.586
church-0.579
Token RE
Token ID4526
Feature activation+2.24e-01

Pos logit contributions
 anonymously+2.525
orr+2.355
oux+2.245
 disse+2.199
church+2.152
Neg logit contributions
advertising-3.757
���-3.607
 Dresden-3.547
education-3.523
peria-3.476
TokenP
Token ID47
Feature activation-5.20e-01

Pos logit contributions
advertising+1.579
 Dresden+1.441
defense+1.395
ary+1.380
education+1.379
Neg logit contributions
 anonymously-1.523
orr-1.449
 disse-1.423
oux-1.418
church-1.360
Token no
Token ID645
Feature activation+6.53e-02

Pos logit contributions
 anonymously+3.690
orr+3.497
oux+3.362
 disse+3.321
church+3.252
Neg logit contributions
advertising-4.119
���-3.832
 Dresden-3.813
education-3.716
defense-3.713
Token internet
Token ID5230
Feature activation-1.94e-01

Pos logit contributions
advertising+0.463
 Dresden+0.431
defense+0.419
education+0.419
���+0.418
Neg logit contributions
 anonymously-0.424
orr-0.410
oux-0.391
church-0.382
 disse-0.380
Token connection
Token ID4637
Feature activation-1.49e-01

Pos logit contributions
 anonymously+1.118
orr+1.068
oux+1.022
 disse+0.988
church+0.987
Neg logit contributions
advertising-1.529
 Dresden-1.435
���-1.423
education-1.401
defense-1.397
Token,
Token ID11
Feature activation-2.55e-01

Pos logit contributions
 anonymously+0.978
orr+0.938
oux+0.902
 disse+0.876
church+0.876
Neg logit contributions
advertising-1.059
 Dresden-0.989
���-0.983
education-0.961
defense-0.959
Token but
Token ID475
Feature activation+5.74e-02

Pos logit contributions
 anonymously+1.577
orr+1.499
oux+1.438
 disse+1.397
church+1.392
Neg logit contributions
advertising-1.910
���-1.796
 Dresden-1.786
education-1.741
defense-1.738
Token Broad
Token ID9765
Feature activation-1.93e+00

Pos logit contributions
advertising+0.382
 Dresden+0.359
ary+0.347
���+0.347
education+0.346
Neg logit contributions
 anonymously-0.393
orr-0.383
oux-0.367
church-0.358
 disse-0.356
Tokenband
Token ID3903
Feature activation-1.88e-01

Pos logit contributions
church+7.207
 anonymously+7.171
orr+6.917
nect+6.416
oux+6.395
Neg logit contributions
���-16.654
 Dresden-15.689
advertising-15.091
ebted-14.796
defense-14.791
Token Connection
Token ID26923
Feature activation+2.12e-01

Pos logit contributions
 anonymously+0.870
orr+0.818
oux+0.774
 disse+0.741
church+0.740
Neg logit contributions
advertising-1.689
 Dresden-1.603
���-1.598
education-1.566
defense-1.565
Token (
Token ID357
Feature activation-1.88e-02

Pos logit contributions
advertising+1.411
 Dresden+1.309
ary+1.290
education+1.279
defense+1.266
Neg logit contributions
 anonymously-1.458
orr-1.416
oux-1.358
church-1.327
 disse-1.308
TokenCo
Token ID7222
Feature activation-2.35e-01

Pos logit contributions
 anonymously+0.132
orr+0.127
oux+0.123
 disse+0.120
church+0.120
Neg logit contributions
advertising-0.124
 Dresden-0.115
���-0.115
education-0.112
defense-0.112
Tokenax
Token ID897
Feature activation+1.90e-01

Pos logit contributions
 anonymously+1.534
orr+1.481
oux+1.423
church+1.381
 disse+1.362
Neg logit contributions
advertising-1.659
���-1.590
 Dresden-1.558
defense-1.502
education-1.498
Token The
Token ID383
Feature activation+4.58e-01

Pos logit contributions
 anonymously+0.645
orr+0.619
oux+0.593
church+0.575
 disse+0.572
Neg logit contributions
advertising-0.757
���-0.711
 Dresden-0.705
education-0.691
peria-0.687
Token property
Token ID3119
Feature activation+1.96e-01

Pos logit contributions
advertising+2.956
 Dresden+2.834
ary+2.710
���+2.699
defense+2.697
Neg logit contributions
 anonymously-3.133
orr-3.026
oux-2.977
church-2.852
 disse-2.819
Token now
Token ID783
Feature activation+2.26e-01

Pos logit contributions
advertising+1.364
 Dresden+1.290
ary+1.254
���+1.254
defense+1.245
Neg logit contributions
 anonymously-1.267
orr-1.229
oux-1.190
church-1.157
 disse-1.139
Token belongs
Token ID14448
Feature activation+2.89e-02

Pos logit contributions
advertising+1.556
 Dresden+1.469
ary+1.425
���+1.421
defense+1.413
Neg logit contributions
 anonymously-1.485
orr-1.448
oux-1.399
church-1.358
 disse-1.337
Token to
Token ID284
Feature activation-3.45e-01

Pos logit contributions
advertising+0.177
 Dresden+0.163
education+0.158
defense+0.157
���+0.157
Neg logit contributions
 anonymously-0.216
orr-0.210
oux-0.203
 disse-0.199
church-0.198
Token Cit
Token ID15792
Feature activation-1.92e+00

Pos logit contributions
 anonymously+2.436
orr+2.329
oux+2.239
church+2.192
 disse+2.191
Neg logit contributions
advertising-2.267
���-2.111
 Dresden-2.076
education-2.041
peria-2.033
Tokenib
Token ID571
Feature activation+5.15e-02

Pos logit contributions
oux+8.030
orr+7.631
 anonymously+7.406
ucci+7.239
church+6.826
Neg logit contributions
���-17.329
 Dresden-15.274
advertising-14.663
 contestant-14.185
lance-14.162
Tokenank
Token ID962
Feature activation-9.57e-01

Pos logit contributions
advertising+0.256
 Dresden+0.229
ary+0.223
education+0.222
defense+0.221
Neg logit contributions
 anonymously-0.449
orr-0.432
oux-0.419
 disse-0.418
church-0.411
Token.
Token ID13
Feature activation-2.96e-02

Pos logit contributions
 anonymously+5.673
orr+5.454
oux+5.198
 disse+4.923
church+4.911
Neg logit contributions
advertising-7.131
���-6.798
 Dresden-6.414
education-6.406
defense-6.359
Token
Token ID198
Feature activation+1.62e-03

Pos logit contributions
 anonymously+0.184
orr+0.177
oux+0.169
church+0.164
 disse+0.164
Neg logit contributions
advertising-0.220
���-0.206
 Dresden-0.205
education-0.200
defense-0.199
Token
Token ID198
Feature activation+1.90e-01

Pos logit contributions
advertising+0.012
 Dresden+0.011
ary+0.011
education+0.011
defense+0.011
Neg logit contributions
 anonymously-0.011
orr-0.010
oux-0.010
 disse-0.010
church-0.009
Token (
Token ID357
Feature activation+1.35e-02

Pos logit contributions
 anonymously+0.324
orr+0.310
oux+0.299
 disse+0.290
church+0.290
Neg logit contributions
advertising-0.357
 Dresden-0.333
���-0.332
education-0.324
defense-0.323
Tokenwhich
Token ID4758
Feature activation+1.07e+00

Pos logit contributions
advertising+0.095
 Dresden+0.088
���+0.087
education+0.086
defense+0.086
Neg logit contributions
 anonymously-0.090
orr-0.086
oux-0.083
 disse-0.081
church-0.080
Token has
Token ID468
Feature activation-1.41e-01

Pos logit contributions
 Dresden+5.641
advertising+5.445
ary+5.325
���+5.098
education+4.991
Neg logit contributions
orr-7.902
 anonymously-7.745
oux-7.668
church-7.407
scribe-7.251
Token the
Token ID262
Feature activation-4.52e-01

Pos logit contributions
 anonymously+0.997
orr+0.960
oux+0.926
church+0.901
 disse+0.899
Neg logit contributions
advertising-0.931
 Dresden-0.868
���-0.861
education-0.839
defense-0.838
Token Hyp
Token ID21209
Feature activation-2.93e-01

Pos logit contributions
 anonymously+2.978
orr+2.818
oux+2.735
 disse+2.696
church+2.649
Neg logit contributions
advertising-3.210
���-2.978
 Dresden-2.907
peria-2.900
education-2.876
Tokenac
Token ID330
Feature activation-1.91e+00

Pos logit contributions
 anonymously+1.879
orr+1.835
oux+1.784
church+1.702
 disse+1.655
Neg logit contributions
advertising-2.091
���-1.990
 Dresden-1.981
peria-1.884
education-1.879
Tokenros
Token ID4951
Feature activation-3.43e-01

Pos logit contributions
orr+5.338
oux+4.960
 anonymously+4.589
sters+4.159
church+4.108
Neg logit contributions
���-19.795
advertising-17.777
 Dresden-17.175
ebted-16.774
 contestant-16.746
Tokenaurus
Token ID22302
Feature activation+2.77e-04

Pos logit contributions
 anonymously+1.869
orr+1.799
oux+1.731
church+1.646
 disse+1.625
Neg logit contributions
advertising-2.772
���-2.694
 Dresden-2.642
education-2.559
defense-2.542
Token egg
Token ID5935
Feature activation-4.00e-01

Pos logit contributions
advertising+0.002
 Dresden+0.002
���+0.002
ary+0.002
education+0.002
Neg logit contributions
 anonymously-0.002
orr-0.002
oux-0.002
church-0.002
 disse-0.002
Token)
Token ID8
Feature activation+3.68e-01

Pos logit contributions
 anonymously+2.832
orr+2.702
oux+2.627
 disse+2.568
church+2.552
Neg logit contributions
advertising-2.636
���-2.472
 Dresden-2.412
peria-2.357
defense-2.349
Token into
Token ID656
Feature activation-7.04e-01

Pos logit contributions
advertising+2.510
 Dresden+2.408
���+2.316
education+2.311
ary+2.302
Neg logit contributions
 anonymously-2.394
orr-2.359
oux-2.256
church-2.197
 disse-2.167
Tokenag
Token ID363
Feature activation+5.68e-02

Pos logit contributions
advertising+0.987
 Dresden+0.906
���+0.894
education+0.892
defense+0.884
Neg logit contributions
 anonymously-1.057
orr-1.009
oux-0.975
 disse-0.962
church-0.953
Tokenression
Token ID2234
Feature activation-7.22e-02

Pos logit contributions
advertising+0.413
 Dresden+0.384
education+0.378
ary+0.377
���+0.376
Neg logit contributions
 anonymously-0.365
orr-0.343
oux-0.329
 disse-0.325
church-0.323
Token de
Token ID390
Feature activation+2.46e-02

Pos logit contributions
 anonymously+0.448
orr+0.428
oux+0.410
church+0.399
 disse+0.398
Neg logit contributions
advertising-0.538
 Dresden-0.505
���-0.501
education-0.491
defense-0.490
Token t
Token ID256
Feature activation-5.10e-01

Pos logit contributions
advertising+0.179
 Dresden+0.169
���+0.167
education+0.163
ary+0.163
Neg logit contributions
 anonymously-0.156
orr-0.149
oux-0.143
church-0.140
 disse-0.138
Tokenout
Token ID448
Feature activation-9.42e-01

Pos logit contributions
 anonymously+3.183
orr+3.123
oux+3.049
 disse+2.887
church+2.885
Neg logit contributions
advertising-3.673
���-3.517
 Dresden-3.483
defense-3.345
education-3.331
Token cit
Token ID15433
Feature activation-1.91e+00

Pos logit contributions
 anonymously+5.810
orr+5.439
oux+5.376
 disse+5.226
church+4.959
Neg logit contributions
advertising-6.699
���-6.406
 Dresden-6.353
defense-6.212
peria-6.133
Tokenoy
Token ID726
Feature activation+2.32e-02

Pos logit contributions
oux+12.058
 anonymously+11.970
orr+11.693
ucci+10.996
 disse+10.952
Neg logit contributions
���-10.809
advertising-10.730
 Dresden-10.193
defense-9.592
eanor-9.443
Tokenen
Token ID268
Feature activation-6.05e-01

Pos logit contributions
advertising+0.172
 Dresden+0.161
���+0.159
education+0.158
ary+0.157
Neg logit contributions
 anonymously-0.144
orr-0.137
oux-0.131
church-0.128
 disse-0.128
Token Imp
Token ID9855
Feature activation-1.89e-01

Pos logit contributions
 anonymously+3.762
orr+3.591
oux+3.495
 disse+3.388
church+3.304
Neg logit contributions
advertising-4.397
���-4.180
 Dresden-4.103
education-4.019
defense-4.017
Tokené
Token ID2634
Feature activation-3.17e-02

Pos logit contributions
 anonymously+1.196
orr+1.150
oux+1.111
church+1.069
 disse+1.069
Neg logit contributions
advertising-1.370
 Dresden-1.286
���-1.284
education-1.248
defense-1.248
Tokenrial
Token ID4454
Feature activation-1.27e-01

Pos logit contributions
 anonymously+0.205
orr+0.195
oux+0.188
 disse+0.183
church+0.183
Neg logit contributions
advertising-0.228
 Dresden-0.213
���-0.211
defense-0.207
education-0.207
Tokenh
Token ID71
Feature activation-3.84e-01

Pos logit contributions
 anonymously+1.530
orr+1.498
oux+1.449
church+1.427
 disse+1.403
Neg logit contributions
advertising-0.893
���-0.836
 Dresden-0.834
peria-0.789
defense-0.770
Tokeniking
Token ID14132
Feature activation-1.08e-01

Pos logit contributions
 anonymously+2.696
orr+2.669
oux+2.573
church+2.474
 disse+2.427
Neg logit contributions
advertising-2.483
���-2.432
 Dresden-2.322
defense-2.237
peria-2.216
Token
Token ID447
Feature activation-1.07e-01

Pos logit contributions
 anonymously+0.738
orr+0.709
oux+0.682
 disse+0.666
church+0.665
Neg logit contributions
advertising-0.732
 Dresden-0.684
���-0.672
education-0.663
ary-0.662
Token
Token ID247
Feature activation-1.09e-01

Pos logit contributions
 anonymously+1.053
orr+1.026
oux+0.997
church+0.980
 disse+0.978
Neg logit contributions
advertising-0.410
 Dresden-0.360
���-0.355
education-0.339
defense-0.337
Tokens
Token ID82
Feature activation-4.75e-01

Pos logit contributions
 anonymously+0.736
orr+0.708
oux+0.681
church+0.662
 disse+0.660
Neg logit contributions
advertising-0.748
���-0.697
 Dresden-0.696
education-0.674
defense-0.672
Token holy
Token ID11386
Feature activation-1.90e+00

Pos logit contributions
 anonymously+3.185
orr+3.060
oux+2.982
church+2.871
 disse+2.856
Neg logit contributions
advertising-3.265
���-3.064
 Dresden-3.012
education-2.935
defense-2.932
Token gra
Token ID7933
Feature activation+1.42e-02

Pos logit contributions
oux+8.857
 anonymously+8.550
church+8.401
orr+8.213
 disse+7.475
Neg logit contributions
���-15.130
advertising-13.512
peria-12.922
ebted-12.909
 Morales-12.825
Tokenil
Token ID346
Feature activation+3.15e-01

Pos logit contributions
advertising+0.093
 Dresden+0.087
���+0.086
education+0.084
defense+0.084
Neg logit contributions
 anonymously-0.101
orr-0.097
oux-0.093
 disse-0.091
church-0.091
Token.
Token ID13
Feature activation-1.04e-01

Pos logit contributions
advertising+2.178
 Dresden+2.044
ary+1.997
education+1.987
defense+1.984
Neg logit contributions
 anonymously-2.066
orr-1.996
oux-1.914
 disse-1.877
church-1.850
Token In
Token ID554
Feature activation-3.27e-01

Pos logit contributions
 anonymously+0.748
orr+0.718
oux+0.695
 disse+0.678
church+0.676
Neg logit contributions
advertising-0.678
 Dresden-0.632
���-0.621
education-0.609
defense-0.607
Token Chicago
Token ID4842
Feature activation-1.07e-01

Pos logit contributions
 anonymously+2.423
orr+2.335
oux+2.260
church+2.204
 disse+2.198
Neg logit contributions
advertising-2.025
���-1.886
 Dresden-1.832
defense-1.803
education-1.801
Token with
Token ID351
Feature activation-5.16e-01

Pos logit contributions
 anonymously+1.854
orr+1.758
oux+1.694
church+1.640
 disse+1.637
Neg logit contributions
advertising-2.128
���-2.009
 Dresden-2.000
education-1.933
peria-1.932
Token him
Token ID683
Feature activation+1.01e-01

Pos logit contributions
 anonymously+3.218
orr+3.048
oux+2.919
 disse+2.859
church+2.828
Neg logit contributions
advertising-3.814
���-3.585
 Dresden-3.577
education-3.476
defense-3.475
Token.
Token ID13
Feature activation-7.65e-02

Pos logit contributions
advertising+0.709
 Dresden+0.662
���+0.649
education+0.643
ary+0.643
Neg logit contributions
 anonymously-0.664
orr-0.644
oux-0.619
church-0.603
 disse-0.602
Token<|endoftext|>
Token ID50256
Feature activation-3.54e-02

Pos logit contributions
 anonymously+0.515
orr+0.494
oux+0.477
 disse+0.465
church+0.463
Neg logit contributions
advertising-0.530
 Dresden-0.495
���-0.487
education-0.479
defense-0.478
TokenS
Token ID50
Feature activation-1.35e-01

Pos logit contributions
 anonymously+0.248
orr+0.238
oux+0.230
 disse+0.224
church+0.224
Neg logit contributions
advertising-0.235
 Dresden-0.219
���-0.217
education-0.212
defense-0.211
TokenAC
Token ID2246
Feature activation-1.90e+00

Pos logit contributions
 anonymously+0.932
orr+0.900
oux+0.868
church+0.841
 disse+0.840
Neg logit contributions
advertising-0.909
���-0.847
 Dresden-0.847
education-0.819
defense-0.819
TokenRAM
Token ID24115
Feature activation-1.21e+00

Pos logit contributions
 anonymously+10.717
orr+10.408
oux+9.901
church+9.851
HT+9.694
Neg logit contributions
advertising-12.664
���-12.161
ebted-11.895
defense-11.490
 Dresden-11.292
TokenENT
Token ID3525
Feature activation-1.18e-02

Pos logit contributions
 anonymously+4.021
orr+3.744
oux+3.341
church+3.240
 disse+3.103
Neg logit contributions
advertising-11.831
���-11.586
education-11.389
defense-11.254
lance-11.217
TokenO
Token ID46
Feature activation-4.88e-01

Pos logit contributions
 anonymously+0.089
orr+0.086
oux+0.083
 disse+0.081
church+0.081
Neg logit contributions
advertising-0.072
 Dresden-0.066
���-0.065
ary-0.064
education-0.064
Token
Token ID851
Feature activation+2.51e-01

Pos logit contributions
 anonymously+3.327
orr+3.143
oux+3.037
church+2.997
 disse+2.938
Neg logit contributions
advertising-3.294
���-3.089
 Dresden-3.051
education-2.967
defense-2.950
Token Thousands
Token ID30405
Feature activation+9.87e-01

Pos logit contributions
advertising+1.581
 Dresden+1.466
defense+1.404
education+1.399
���+1.396
Neg logit contributions
 anonymously-1.842
orr-1.794
oux-1.736
 disse-1.702
church-1.687
Token://
Token ID1378
Feature activation-1.17e+00

Pos logit contributions
 anonymously+3.516
orr+3.443
oux+3.207
church+3.076
nect+2.887
Neg logit contributions
���-5.720
advertising-5.197
 Dresden-5.088
peria-4.984
ebted-4.963
Tokent
Token ID83
Feature activation-9.87e-01

Pos logit contributions
orr+4.267
 anonymously+4.164
oux+3.908
church+3.725
nect+3.465
Neg logit contributions
���-11.994
 Dresden-10.682
advertising-10.424
peria-10.205
ebted-10.174
Token.
Token ID13
Feature activation-1.27e+00

Pos logit contributions
orr+4.831
 anonymously+4.815
oux+4.559
church+4.348
ucci+4.045
Neg logit contributions
���-9.342
advertising-7.734
 Dresden-7.650
peria-7.496
ebted-7.384
Tokenco
Token ID1073
Feature activation-6.93e-01

Pos logit contributions
orr+5.993
 anonymously+5.630
oux+5.479
church+5.295
ste+5.034
Neg logit contributions
���-12.159
 Dresden-10.244
advertising-10.043
amily-9.855
ebted-9.746
Token/
Token ID14
Feature activation-1.05e+00

Pos logit contributions
 anonymously+3.796
orr+3.717
oux+3.470
church+3.372
nect+3.154
Neg logit contributions
���-5.893
advertising-5.339
 Dresden-5.122
peria-5.059
amily-5.045
Tokenl
Token ID75
Feature activation-1.88e+00

Pos logit contributions
orr+6.806
 anonymously+6.546
oux+6.534
church+6.324
kef+5.984
Neg logit contributions
���-6.997
advertising-6.784
 Dresden-6.550
education-6.150
peria-6.126
Tokeng
Token ID70
Feature activation-1.46e+00

Pos logit contributions
orr+10.961
oux+10.586
church+10.270
HT+10.083
 anonymously+10.003
Neg logit contributions
���-12.379
advertising-10.948
 Dresden-10.936
ebted-10.313
education-10.098
TokenS
Token ID50
Feature activation-7.93e-01

Pos logit contributions
orr+8.888
oux+8.646
church+8.317
 anonymously+8.315
HT+8.002
Neg logit contributions
���-9.890
advertising-9.119
 Dresden-8.930
ebted-8.499
education-8.402
Tokenwh
Token ID1929
Feature activation-9.69e-01

Pos logit contributions
orr+5.126
 anonymously+5.031
oux+4.973
church+4.775
 disse+4.563
Neg logit contributions
���-5.319
advertising-5.314
 Dresden-5.027
education-4.786
ebted-4.780
TokenX
Token ID55
Feature activation-3.41e-01

Pos logit contributions
orr+6.270
 anonymously+6.083
oux+6.032
church+5.842
kef+5.511
Neg logit contributions
���-6.461
advertising-6.321
 Dresden-6.075
education-5.739
ebted-5.719
TokenUm
Token ID37280
Feature activation-5.37e-01

Pos logit contributions
 anonymously+2.308
orr+2.258
oux+2.177
church+2.112
 disse+2.079
Neg logit contributions
advertising-2.297
���-2.186
 Dresden-2.159
education-2.077
defense-2.071
Token through
Token ID832
Feature activation-1.02e+00

Pos logit contributions
 anonymously+6.034
orr+5.873
church+5.486
oux+5.421
nect+5.300
Neg logit contributions
advertising-7.990
���-7.505
 Dresden-7.434
peria-7.238
ebted-7.196
Token a
Token ID257
Feature activation-8.10e-01

Pos logit contributions
 anonymously+7.306
orr+6.863
 disse+6.596
oux+6.560
church+6.417
Neg logit contributions
advertising-6.469
���-6.063
 Dresden-5.808
peria-5.769
ebted-5.690
Token ridiculous
Token ID11441
Feature activation-7.98e-01

Pos logit contributions
 anonymously+4.975
orr+4.614
 disse+4.412
oux+4.392
church+4.292
Neg logit contributions
advertising-6.076
���-5.659
 Dresden-5.493
peria-5.437
education-5.426
Token and
Token ID290
Feature activation-6.64e-01

Pos logit contributions
 anonymously+4.893
orr+4.507
 disse+4.341
oux+4.333
church+4.220
Neg logit contributions
advertising-5.941
���-5.664
 Dresden-5.424
peria-5.384
ebted-5.339
Token unpopular
Token ID24166
Feature activation-1.31e+00

Pos logit contributions
 anonymously+4.094
orr+3.763
 disse+3.597
oux+3.594
church+3.471
Neg logit contributions
advertising-4.982
���-4.666
 Dresden-4.597
peria-4.585
defense-4.533
Token bill
Token ID2855
Feature activation-1.88e+00

Pos logit contributions
 anonymously+7.141
orr+6.375
 disse+6.356
oux+6.213
church+5.968
Neg logit contributions
advertising-10.044
���-9.758
peria-9.154
 Dresden-9.086
lance-9.019
Token in
Token ID287
Feature activation-1.12e+00

Pos logit contributions
 anonymously+10.444
orr+9.026
 disse+8.544
oux+8.223
church+8.148
Neg logit contributions
���-13.461
advertising-13.144
peria-12.528
eanor-12.239
amily-12.224
Token just
Token ID655
Feature activation-8.97e-01

Pos logit contributions
 anonymously+7.519
orr+6.902
oux+6.712
church+6.605
 disse+6.534
Neg logit contributions
advertising-7.524
���-7.289
peria-6.652
education-6.624
ebted-6.622
Token two
Token ID734
Feature activation-2.53e-01

Pos logit contributions
 anonymously+6.293
orr+5.808
oux+5.518
 disse+5.448
church+5.432
Neg logit contributions
advertising-5.827
���-5.535
ebted-5.240
 Dresden-5.230
education-5.227
Token weeks
Token ID2745
Feature activation-1.29e+00

Pos logit contributions
 anonymously+1.661
orr+1.580
oux+1.518
 disse+1.487
church+1.478
Neg logit contributions
advertising-1.791
���-1.682
 Dresden-1.674
education-1.631
defense-1.622
Token?
Token ID30
Feature activation-3.10e-01

Pos logit contributions
 anonymously+6.470
orr+6.067
oux+5.361
 disse+5.302
church+5.174
Neg logit contributions
���-10.367
advertising-10.137
peria-9.734
 Dresden-9.691
amily-9.434
Token,
Token ID11
Feature activation+4.59e-01

Pos logit contributions
 anonymously+1.317
orr+1.259
oux+1.210
 disse+1.177
church+1.174
Neg logit contributions
advertising-1.470
 Dresden-1.372
���-1.368
education-1.336
defense-1.333
Token above
Token ID2029
Feature activation+5.68e-01

Pos logit contributions
advertising+2.900
 Dresden+2.742
ary+2.676
defense+2.610
education+2.606
Neg logit contributions
 anonymously-3.127
orr-3.102
oux-3.069
church-2.966
 disse-2.907
Token all
Token ID477
Feature activation-2.33e-01

Pos logit contributions
advertising+2.925
 Dresden+2.636
education+2.548
ary+2.546
���+2.519
Neg logit contributions
 anonymously-4.707
orr-4.567
oux-4.533
 disse-4.382
church-4.364
Token,
Token ID11
Feature activation-1.56e-01

Pos logit contributions
 anonymously+1.279
orr+1.216
oux+1.160
 disse+1.119
church+1.118
Neg logit contributions
advertising-1.899
 Dresden-1.791
���-1.785
education-1.747
defense-1.744
Token an
Token ID281
Feature activation-2.76e-01

Pos logit contributions
 anonymously+1.035
orr+0.991
oux+0.954
 disse+0.928
church+0.926
Neg logit contributions
advertising-1.098
 Dresden-1.024
���-1.020
education-0.995
defense-0.993
Token uncons
Token ID13962
Feature activation-1.86e+00

Pos logit contributions
 anonymously+1.832
orr+1.742
oux+1.681
 disse+1.632
church+1.623
Neg logit contributions
advertising-1.934
 Dresden-1.806
���-1.796
peria-1.759
defense-1.758
Tokenc
Token ID66
Feature activation-9.77e-01

Pos logit contributions
orr+7.844
 anonymously+7.286
oux+7.112
ucci+6.891
nect+6.887
Neg logit contributions
���-15.233
 Dresden-14.408
advertising-14.098
peria-13.957
lance-13.572
Tokenion
Token ID295
Feature activation-6.41e-01

Pos logit contributions
 anonymously+4.183
orr+4.006
oux+3.865
ococ+3.547
church+3.527
Neg logit contributions
���-9.165
advertising-8.457
 Dresden-8.265
peria-8.094
 Morales-8.013
Tokenable
Token ID540
Feature activation-2.33e-01

Pos logit contributions
 anonymously+3.639
orr+3.550
oux+3.408
church+3.249
 disse+3.182
Neg logit contributions
advertising-4.873
���-4.789
 Dresden-4.690
defense-4.517
peria-4.501
Token gift
Token ID6979
Feature activation-7.32e-01

Pos logit contributions
 anonymously+1.513
orr+1.445
oux+1.389
 disse+1.355
church+1.349
Neg logit contributions
advertising-1.672
���-1.562
 Dresden-1.559
education-1.519
defense-1.515
Token to
Token ID284
Feature activation-4.56e-01

Pos logit contributions
 anonymously+4.685
orr+4.472
oux+4.176
church+4.133
 disse+4.095
Neg logit contributions
advertising-5.221
���-4.974
 Dresden-4.869
peria-4.815
education-4.744
Token flavor
Token ID9565
Feature activation-3.22e-01

Pos logit contributions
 anonymously+1.585
orr+1.519
oux+1.462
church+1.418
 disse+1.417
Neg logit contributions
advertising-1.739
 Dresden-1.631
���-1.613
education-1.580
defense-1.579
Token begins
Token ID6140
Feature activation+3.44e-01

Pos logit contributions
 anonymously+2.093
orr+2.001
oux+1.923
 disse+1.873
church+1.861
Neg logit contributions
advertising-2.300
���-2.161
 Dresden-2.147
defense-2.085
peria-2.083
Token with
Token ID351
Feature activation+8.35e-02

Pos logit contributions
advertising+2.464
 Dresden+2.339
ary+2.284
���+2.265
defense+2.254
Neg logit contributions
 anonymously-2.127
orr-2.081
oux-1.989
church-1.945
 disse-1.922
Token dried
Token ID16577
Feature activation+5.92e-01

Pos logit contributions
advertising+0.486
 Dresden+0.452
���+0.442
defense+0.432
education+0.432
Neg logit contributions
 anonymously-0.650
orr-0.626
oux-0.608
church-0.592
 disse-0.591
Token apr
Token ID46593
Feature activation-4.54e-01

Pos logit contributions
advertising+3.907
 Dresden+3.780
ary+3.643
education+3.551
defense+3.534
Neg logit contributions
 anonymously-3.944
orr-3.750
oux-3.658
church-3.487
 disse-3.465
Tokenic
Token ID291
Feature activation-1.86e+00

Pos logit contributions
 anonymously+3.447
orr+3.422
oux+3.281
church+3.118
 disse+3.079
Neg logit contributions
advertising-2.667
���-2.557
 Dresden-2.508
education-2.402
defense-2.371
Tokenot
Token ID313
Feature activation-6.53e-01

Pos logit contributions
oux+5.248
orr+5.246
 anonymously+4.873
church+4.107
ococ+4.064
Neg logit contributions
advertising-17.759
 Dresden-17.272
lance-16.791
eanor-16.672
���-16.644
Token,
Token ID11
Feature activation-3.99e-01

Pos logit contributions
 anonymously+3.856
orr+3.718
oux+3.525
church+3.424
 disse+3.411
Neg logit contributions
advertising-4.988
���-4.760
 Dresden-4.646
peria-4.601
defense-4.567
Token banana
Token ID25996
Feature activation-8.80e-01

Pos logit contributions
 anonymously+2.507
orr+2.388
oux+2.289
 disse+2.229
church+2.220
Neg logit contributions
advertising-2.941
���-2.774
 Dresden-2.723
education-2.671
peria-2.670
Token,
Token ID11
Feature activation-1.83e-01

Pos logit contributions
 anonymously+4.428
orr+4.402
oux+4.023
 disse+3.957
church+3.900
Neg logit contributions
advertising-7.288
���-6.985
 Dresden-6.777
peria-6.752
education-6.737
Token and
Token ID290
Feature activation-4.54e-02

Pos logit contributions
 anonymously+1.146
orr+1.099
oux+1.052
church+1.023
 disse+1.022
Neg logit contributions
advertising-1.358
���-1.271
 Dresden-1.266
education-1.235
defense-1.234
Token might
Token ID1244
Feature activation+2.79e-01

Pos logit contributions
advertising+2.970
ary+2.770
 Dresden+2.764
education+2.689
���+2.654
Neg logit contributions
 anonymously-2.733
orr-2.696
oux-2.606
church-2.489
 disse-2.482
Token become
Token ID1716
Feature activation-2.15e-01

Pos logit contributions
advertising+2.050
 Dresden+1.966
ary+1.949
defense+1.897
education+1.890
Neg logit contributions
 anonymously-1.616
orr-1.605
oux-1.544
church-1.541
 disse-1.462
Token worthless
Token ID28063
Feature activation-1.24e-01

Pos logit contributions
 anonymously+1.457
orr+1.393
oux+1.340
 disse+1.307
church+1.300
Neg logit contributions
advertising-1.488
 Dresden-1.383
���-1.383
education-1.348
defense-1.343
Token.
Token ID13
Feature activation-2.52e-01

Pos logit contributions
 anonymously+0.798
orr+0.761
oux+0.732
 disse+0.711
church+0.709
Neg logit contributions
advertising-0.896
���-0.841
 Dresden-0.838
education-0.816
defense-0.814
Token The
Token ID383
Feature activation-2.83e-01

Pos logit contributions
 anonymously+1.745
orr+1.675
oux+1.611
church+1.569
 disse+1.562
Neg logit contributions
advertising-1.689
���-1.593
 Dresden-1.568
education-1.530
defense-1.528
Token gap
Token ID7625
Feature activation-1.85e+00

Pos logit contributions
 anonymously+1.810
orr+1.723
oux+1.656
 disse+1.613
church+1.609
Neg logit contributions
advertising-2.059
���-1.926
 Dresden-1.912
education-1.869
defense-1.868
Token was
Token ID373
Feature activation-1.01e+00

Pos logit contributions
 anonymously+8.893
orr+8.041
oux+7.477
church+7.329
 disse+7.328
Neg logit contributions
���-13.933
advertising-13.818
amily-12.782
defense-12.766
peria-12.742
Token filled
Token ID5901
Feature activation-4.01e-01

Pos logit contributions
 anonymously+5.624
orr+5.182
oux+4.989
 disse+4.869
church+4.720
Neg logit contributions
advertising-8.110
���-7.638
 Dresden-7.457
peria-7.391
defense-7.382
Token by
Token ID416
Feature activation-1.96e-01

Pos logit contributions
 anonymously+2.849
orr+2.692
oux+2.615
church+2.538
 disse+2.537
Neg logit contributions
advertising-2.645
���-2.482
 Dresden-2.433
defense-2.383
education-2.373
Token,
Token ID11
Feature activation-4.88e-01

Pos logit contributions
 anonymously+1.379
orr+1.320
oux+1.270
 disse+1.240
church+1.236
Neg logit contributions
advertising-1.307
���-1.218
 Dresden-1.209
education-1.177
defense-1.176
Token among
Token ID1871
Feature activation-6.48e-01

Pos logit contributions
 anonymously+2.940
orr+2.749
oux+2.604
 disse+2.526
church+2.514
Neg logit contributions
advertising-3.750
���-3.591
 Dresden-3.516
peria-3.466
education-3.428
Token
Token ID198
Feature activation+4.49e-01

Pos logit contributions
advertising+1.263
 Dresden+1.191
ary+1.140
education+1.137
defense+1.133
Neg logit contributions
 anonymously-1.298
orr-1.244
oux-1.207
 disse-1.174
church-1.169
Token
Token ID198
Feature activation+5.11e-01

Pos logit contributions
advertising+2.911
ary+2.632
 Dresden+2.629
education+2.541
defense+2.536
Neg logit contributions
 anonymously-3.240
 disse-3.092
orr-3.078
oux-3.011
church-2.923
TokenBe
Token ID3856
Feature activation+1.10e+00

Pos logit contributions
advertising+3.184
 Dresden+2.758
ary+2.756
education+2.748
defense+2.728
Neg logit contributions
 anonymously-3.956
orr-3.730
oux-3.659
 disse-3.635
church-3.482
Token that
Token ID326
Feature activation-1.40e-01

Pos logit contributions
advertising+6.912
���+6.487
 Dresden+6.469
ary+6.411
education+6.354
Neg logit contributions
 anonymously-7.336
orr-6.514
oux-6.493
 disse-6.419
church-6.382
Token as
Token ID355
Feature activation-1.45e+00

Pos logit contributions
 anonymously+1.116
orr+1.077
oux+1.044
 disse+1.020
church+1.019
Neg logit contributions
advertising-0.798
 Dresden-0.733
���-0.728
education-0.706
defense-0.704
Token it
Token ID340
Feature activation-1.85e+00

Pos logit contributions
 anonymously+6.106
orr+5.693
oux+5.036
church+4.700
HT+4.620
Neg logit contributions
���-13.025
ebted-12.288
advertising-12.253
amily-12.009
 Dresden-11.689
Token may
Token ID743
Feature activation-1.22e+00

Pos logit contributions
 anonymously+8.137
orr+7.112
oux+6.261
ucci+6.258
 then+6.104
Neg logit contributions
���-16.810
peria-14.876
advertising-14.648
ebted-14.472
amily-14.160
Token,
Token ID11
Feature activation+5.65e-01

Pos logit contributions
 anonymously+5.193
orr+4.727
oux+4.407
 disse+4.297
church+4.090
Neg logit contributions
advertising-10.586
peria-9.943
���-9.916
 Dresden-9.904
defense-9.892
Token there
Token ID612
Feature activation+3.15e-01

Pos logit contributions
advertising+3.474
 Dresden+3.295
ary+3.122
���+3.118
education+3.110
Neg logit contributions
 anonymously-4.059
orr-3.992
oux-3.830
church-3.753
nect-3.627
Token seems
Token ID2331
Feature activation+2.61e-01

Pos logit contributions
advertising+2.169
 Dresden+2.004
���+1.972
ary+1.969
education+1.958
Neg logit contributions
 anonymously-2.098
orr-2.027
oux-1.969
church-1.909
 disse-1.897
Token to
Token ID284
Feature activation-5.96e-01

Pos logit contributions
advertising+1.843
 Dresden+1.736
���+1.704
ary+1.684
education+1.676
Neg logit contributions
 anonymously-1.696
orr-1.645
oux-1.580
church-1.536
 disse-1.528
Token
Token ID198
Feature activation+2.42e-01

Pos logit contributions
 anonymously+1.182
orr+1.137
oux+1.096
 disse+1.066
church+1.066
Neg logit contributions
advertising-1.132
 Dresden-1.050
���-1.048
education-1.020
defense-1.018
Token
Token ID198
Feature activation+1.89e-01

Pos logit contributions
advertising+1.644
 Dresden+1.507
ary+1.498
education+1.464
defense+1.463
Neg logit contributions
 anonymously-1.676
orr-1.592
 disse-1.565
oux-1.548
church-1.504
TokenIf
Token ID1532
Feature activation+8.11e-01

Pos logit contributions
advertising+1.219
 Dresden+1.108
ary+1.089
defense+1.087
education+1.086
Neg logit contributions
 anonymously-1.385
orr-1.313
oux-1.280
 disse-1.257
church-1.232
Token we
Token ID356
Feature activation+4.26e-01

Pos logit contributions
advertising+4.844
 Dresden+4.691
ary+4.582
defense+4.397
���+4.388
Neg logit contributions
 anonymously-5.779
orr-5.587
oux-5.473
church-5.321
 disse-5.297
Token simply
Token ID2391
Feature activation+2.39e-01

Pos logit contributions
advertising+3.016
 Dresden+2.884
ary+2.836
���+2.823
defense+2.766
Neg logit contributions
 anonymously-2.627
orr-2.499
oux-2.487
church-2.464
 disse-2.325
Token want
Token ID765
Feature activation+1.35e+00

Pos logit contributions
advertising+1.663
 Dresden+1.600
ary+1.560
���+1.548
defense+1.525
Neg logit contributions
 anonymously-1.515
orr-1.470
oux-1.443
church-1.415
 disse-1.356
Token to
Token ID284
Feature activation+4.66e-01

Pos logit contributions
ary+6.856
advertising+6.799
 Dresden+6.681
���+6.364
defense+6.253
Neg logit contributions
 anonymously-9.588
oux-9.392
orr-9.360
church-9.055
 disse-9.014
Token write
Token ID3551
Feature activation+9.13e-01

Pos logit contributions
advertising+2.855
 Dresden+2.783
ary+2.767
defense+2.621
���+2.598
Neg logit contributions
 anonymously-3.127
orr-3.113
oux-3.095
church-3.011
 disse-2.904
Token an
Token ID281
Feature activation-8.78e-03

Pos logit contributions
advertising+5.046
 Dresden+5.027
ary+4.849
���+4.688
defense+4.686
Neg logit contributions
orr-6.380
 anonymously-6.347
oux-6.327
church-6.103
kef-5.890
Token Own
Token ID11744
Feature activation-1.88e-01

Pos logit contributions
 anonymously+0.057
orr+0.055
oux+0.053
church+0.051
 disse+0.051
Neg logit contributions
advertising-0.063
 Dresden-0.059
���-0.058
education-0.057
ary-0.057
Tokened
Token ID276
Feature activation-1.53e-01

Pos logit contributions
 anonymously+1.516
orr+1.457
oux+1.411
 disse+1.382
church+1.376
Neg logit contributions
advertising-1.048
���-0.964
 Dresden-0.958
education-0.924
defense-0.922
Token start
Token ID923
Feature activation-1.75e-01

Pos logit contributions
 anonymously+1.837
orr+1.740
oux+1.646
 disse+1.601
church+1.561
Neg logit contributions
advertising-3.453
���-3.313
 Dresden-3.259
education-3.197
peria-3.196
Token,
Token ID11
Feature activation+3.27e-01

Pos logit contributions
 anonymously+1.138
orr+1.091
oux+1.050
 disse+1.019
church+1.019
Neg logit contributions
advertising-1.244
 Dresden-1.163
���-1.154
education-1.129
defense-1.127
Token after
Token ID706
Feature activation+7.05e-02

Pos logit contributions
advertising+2.418
 Dresden+2.302
defense+2.215
ary+2.208
education+2.189
Neg logit contributions
 anonymously-1.958
orr-1.928
oux-1.865
church-1.796
 disse-1.761
Token Ibrahim
Token ID30917
Feature activation+2.93e-01

Pos logit contributions
advertising+0.476
 Dresden+0.446
���+0.436
defense+0.431
education+0.429
Neg logit contributions
 anonymously-0.482
orr-0.468
oux-0.450
church-0.438
 disse-0.436
Token's
Token ID338
Feature activation+1.69e-01

Pos logit contributions
advertising+1.939
 Dresden+1.853
���+1.786
ary+1.786
defense+1.757
Neg logit contributions
 anonymously-1.948
orr-1.929
oux-1.855
church-1.837
 disse-1.789
Token lawyers
Token ID9326
Feature activation+1.23e+00

Pos logit contributions
advertising+1.165
 Dresden+1.096
���+1.063
defense+1.061
education+1.052
Neg logit contributions
 anonymously-1.114
orr-1.090
oux-1.047
church-1.015
 disse-1.007
Token informed
Token ID7981
Feature activation-1.20e-01

Pos logit contributions
advertising+7.190
 Dresden+7.173
ary+6.827
defense+6.592
 Morales+6.525
Neg logit contributions
orr-7.810
church-7.744
oux-7.503
 anonymously-7.332
nect-7.197
Token the
Token ID262
Feature activation+1.26e-01

Pos logit contributions
 anonymously+0.830
orr+0.797
oux+0.769
 disse+0.748
church+0.747
Neg logit contributions
advertising-0.806
 Dresden-0.750
���-0.746
education-0.727
defense-0.726
Token DOJ
Token ID27055
Feature activation-5.74e-01

Pos logit contributions
advertising+0.827
 Dresden+0.786
���+0.760
defense+0.754
education+0.746
Neg logit contributions
 anonymously-0.884
orr-0.861
oux-0.827
 disse-0.803
church-0.801
Token that
Token ID326
Feature activation-2.07e-01

Pos logit contributions
 anonymously+3.540
orr+3.326
oux+3.182
 disse+3.097
church+3.059
Neg logit contributions
advertising-4.273
���-4.012
 Dresden-3.953
education-3.922
ebted-3.867
Token they
Token ID484
Feature activation-5.08e-02

Pos logit contributions
 anonymously+1.337
orr+1.279
oux+1.230
 disse+1.195
church+1.193
Neg logit contributions
advertising-1.485
 Dresden-1.388
���-1.383
education-1.351
defense-1.346
Token.
Token ID13
Feature activation+6.39e-01

Pos logit contributions
advertising+4.348
ary+3.909
 Dresden+3.772
education+3.720
defense+3.690
Neg logit contributions
 anonymously-5.502
 disse-5.352
orr-5.261
oux-5.170
church-5.052
TokenS
Token ID50
Feature activation-5.95e-02

Pos logit contributions
advertising+4.138
 Dresden+3.672
ary+3.579
defense+3.504
education+3.464
Neg logit contributions
 anonymously-4.653
 disse-4.500
orr-4.392
oux-4.374
church-4.219
Token.,
Token ID1539
Feature activation+9.11e-01

Pos logit contributions
 anonymously+0.409
orr+0.391
oux+0.377
 disse+0.364
church+0.364
Neg logit contributions
advertising-0.406
 Dresden-0.377
���-0.377
education-0.365
defense-0.365
Token also
Token ID635
Feature activation+6.34e-01

Pos logit contributions
advertising+5.595
 Dresden+5.533
ary+5.255
defense+5.147
education+5.060
Neg logit contributions
orr-5.890
 anonymously-5.765
oux-5.750
church-5.644
 disse-5.491
Token dropped
Token ID5710
Feature activation+5.48e-01

Pos logit contributions
advertising+4.061
 Dresden+3.936
ary+3.887
defense+3.718
���+3.703
Neg logit contributions
orr-4.087
 anonymously-4.030
oux-4.015
church-3.981
ococ-3.795
Token nearly
Token ID3016
Feature activation+1.40e+00

Pos logit contributions
advertising+3.379
 Dresden+3.257
ary+3.172
defense+3.039
education+2.969
Neg logit contributions
orr-3.813
 anonymously-3.716
oux-3.691
church-3.579
 disse-3.543
Token 3
Token ID513
Feature activation+5.80e-01

Pos logit contributions
advertising+6.651
 Dresden+6.641
ary+6.560
defense+5.921
 Morales+5.915
Neg logit contributions
orr-10.487
oux-10.228
ucci-10.040
church-9.972
ococ-9.947
Token percent
Token ID1411
Feature activation+6.70e-01

Pos logit contributions
advertising+2.917
 Dresden+2.658
ary+2.633
defense+2.575
education+2.530
Neg logit contributions
 anonymously-4.722
orr-4.666
oux-4.609
 disse-4.505
church-4.497
Token on
Token ID319
Feature activation+4.96e-02

Pos logit contributions
advertising+4.163
 Dresden+3.924
ary+3.886
defense+3.795
peria+3.723
Neg logit contributions
orr-4.610
 anonymously-4.591
oux-4.467
 disse-4.376
church-4.301
Token the
Token ID262
Feature activation-6.87e-02

Pos logit contributions
advertising+0.347
 Dresden+0.325
���+0.319
defense+0.314
education+0.314
Neg logit contributions
 anonymously-0.329
orr-0.317
oux-0.306
church-0.296
 disse-0.295
Token news
Token ID1705
Feature activation+3.87e-01

Pos logit contributions
 anonymously+0.434
orr+0.417
oux+0.402
church+0.390
 disse+0.389
Neg logit contributions
advertising-0.501
 Dresden-0.470
���-0.465
education-0.456
defense-0.455
Token need
Token ID761
Feature activation+2.06e-01

Pos logit contributions
 anonymously+3.192
orr+3.037
oux+2.912
 disse+2.854
church+2.817
Neg logit contributions
advertising-3.555
���-3.323
 Dresden-3.297
education-3.226
defense-3.210
Token it
Token ID340
Feature activation+1.35e-01

Pos logit contributions
advertising+1.418
 Dresden+1.323
���+1.306
defense+1.287
ary+1.284
Neg logit contributions
 anonymously-1.382
orr-1.327
oux-1.275
church-1.247
 disse-1.240
Token.
Token ID13
Feature activation-2.59e-01

Pos logit contributions
advertising+0.920
 Dresden+0.861
���+0.843
ary+0.834
education+0.833
Neg logit contributions
 anonymously-0.905
orr-0.871
oux-0.841
church-0.822
 disse-0.817
Token
Token ID198
Feature activation-6.86e-02

Pos logit contributions
 anonymously+1.762
orr+1.692
oux+1.630
 disse+1.585
church+1.584
Neg logit contributions
advertising-1.770
 Dresden-1.648
���-1.641
education-1.600
defense-1.597
Token
Token ID198
Feature activation-7.56e-02

Pos logit contributions
 anonymously+0.441
orr+0.420
oux+0.405
 disse+0.398
church+0.393
Neg logit contributions
advertising-0.499
 Dresden-0.465
education-0.452
ary-0.452
defense-0.451
TokenRA
Token ID3861
Feature activation+1.10e+00

Pos logit contributions
 anonymously+0.524
orr+0.503
oux+0.485
 disse+0.473
church+0.471
Neg logit contributions
advertising-0.509
 Dresden-0.472
���-0.468
education-0.459
defense-0.458
TokenID
Token ID2389
Feature activation+3.58e-01

Pos logit contributions
advertising+6.558
defense+6.003
 Dresden+5.975
education+5.966
peria+5.871
Neg logit contributions
 anonymously-7.417
orr-7.255
oux-6.846
 disse-6.802
church-6.781
Token4
Token ID19
Feature activation-4.54e-01

Pos logit contributions
advertising+2.477
 Dresden+2.334
���+2.277
defense+2.273
education+2.261
Neg logit contributions
 anonymously-2.328
orr-2.216
oux-2.135
church-2.106
 disse-2.084
Token on
Token ID319
Feature activation+8.77e-02

Pos logit contributions
 anonymously+3.042
orr+2.920
oux+2.819
 disse+2.736
church+2.721
Neg logit contributions
advertising-3.132
���-2.926
 Dresden-2.904
education-2.822
defense-2.810
Token tape
Token ID9154
Feature activation+6.17e-02

Pos logit contributions
advertising+0.606
 Dresden+0.572
���+0.560
education+0.553
ary+0.551
Neg logit contributions
 anonymously-0.583
orr-0.561
oux-0.539
church-0.526
 disse-0.522
Token.
Token ID764
Feature activation+4.09e-01

Pos logit contributions
advertising+0.420
 Dresden+0.391
���+0.386
education+0.380
defense+0.380
Neg logit contributions
 anonymously-0.420
orr-0.403
oux-0.388
church-0.381
 disse-0.377
TokenThe
Token ID464
Feature activation+4.05e-01

Pos logit contributions
advertising+2.065
 Dresden+1.843
education+1.827
ary+1.815
defense+1.810
Neg logit contributions
 anonymously-2.334
orr-2.235
oux-2.180
 disse-2.173
church-2.080
Token flyer
Token ID42464
Feature activation+5.70e-01

Pos logit contributions
advertising+2.877
 Dresden+2.774
ary+2.641
���+2.636
education+2.626
Neg logit contributions
 anonymously-2.488
orr-2.470
oux-2.374
 disse-2.309
church-2.291
Token shows
Token ID2523
Feature activation+6.77e-01

Pos logit contributions
advertising+3.785
 Dresden+3.677
ary+3.581
education+3.461
defense+3.407
Neg logit contributions
orr-3.633
 anonymously-3.559
oux-3.509
church-3.429
 disse-3.345
Token a
Token ID257
Feature activation+4.31e-01

Pos logit contributions
advertising+3.901
 Dresden+3.801
ary+3.510
education+3.486
���+3.450
Neg logit contributions
orr-5.027
 anonymously-4.941
oux-4.847
church-4.623
 disse-4.598
Token newspaper
Token ID7533
Feature activation+1.24e+00

Pos logit contributions
advertising+2.872
 Dresden+2.796
���+2.664
ary+2.639
education+2.614
Neg logit contributions
orr-2.827
 anonymously-2.826
oux-2.709
church-2.585
 disse-2.585
Token headline
Token ID16534
Feature activation+1.25e+00

Pos logit contributions
advertising+7.114
 Dresden+6.880
ary+6.849
 contestant+6.649
education+6.512
Neg logit contributions
orr-8.057
oux-7.915
 anonymously-7.543
kef-7.467
church-7.410
Token taken
Token ID2077
Feature activation+8.74e-01

Pos logit contributions
advertising+7.244
 Dresden+7.242
ary+7.183
 Morales+6.682
education+6.600
Neg logit contributions
orr-7.972
oux-7.892
church-7.442
nect-7.435
 anonymously-7.393
Token from
Token ID422
Feature activation+4.69e-01

Pos logit contributions
advertising+5.036
 Dresden+4.968
ary+4.798
education+4.547
defense+4.531
Neg logit contributions
orr-6.298
oux-5.868
church-5.856
nect-5.782
 anonymously-5.732
Token the
Token ID262
Feature activation+3.72e-01

Pos logit contributions
advertising+2.624
 Dresden+2.469
���+2.340
education+2.326
ary+2.304
Neg logit contributions
 anonymously-3.634
orr-3.611
oux-3.488
 disse-3.406
church-3.391
Token National
Token ID2351
Feature activation+2.37e-01

Pos logit contributions
advertising+2.478
 Dresden+2.383
���+2.259
education+2.256
ary+2.248
Neg logit contributions
 anonymously-2.469
orr-2.462
oux-2.358
 disse-2.305
church-2.273
Token Post
Token ID2947
Feature activation+4.62e-01

Pos logit contributions
advertising+1.798
 Dresden+1.730
ary+1.683
education+1.662
���+1.656
Neg logit contributions
 anonymously-1.359
orr-1.334
oux-1.287
 disse-1.256
church-1.216
Token
Token ID198
Feature activation-2.09e-01

Pos logit contributions
 anonymously+0.001
orr+0.001
oux+0.001
 disse+0.001
church+0.001
Neg logit contributions
advertising-0.002
 Dresden-0.002
���-0.002
ary-0.002
education-0.002
Token
Token ID198
Feature activation+1.18e-01

Pos logit contributions
 anonymously+1.289
orr+1.234
oux+1.182
church+1.146
 disse+1.141
Neg logit contributions
advertising-1.556
���-1.472
 Dresden-1.461
education-1.421
defense-1.419
TokenMaterials
Token ID41657
Feature activation+4.51e-01

Pos logit contributions
advertising+0.774
 Dresden+0.705
education+0.689
defense+0.688
ary+0.685
Neg logit contributions
 anonymously-0.857
orr-0.818
oux-0.793
 disse-0.783
church-0.767
Token
Token ID198
Feature activation+4.11e-01

Pos logit contributions
advertising+3.184
 Dresden+3.003
education+2.924
ary+2.911
���+2.899
Neg logit contributions
 anonymously-2.879
orr-2.779
oux-2.659
church-2.567
 disse-2.551
Token
Token ID198
Feature activation+7.47e-01

Pos logit contributions
advertising+2.686
 Dresden+2.442
ary+2.438
education+2.374
defense+2.366
Neg logit contributions
 anonymously-2.938
 disse-2.809
orr-2.797
oux-2.735
church-2.644
Token.
Token ID13
Feature activation+7.57e-01

Pos logit contributions
advertising+4.748
education+4.230
defense+4.196
ary+4.188
 Dresden+4.185
Neg logit contributions
 anonymously-5.472
orr-5.144
 disse-5.067
oux-5.009
church-4.792
Token
Token ID198
Feature activation+9.86e-02

Pos logit contributions
advertising+5.118
 Dresden+4.853
defense+4.714
ary+4.706
���+4.702
Neg logit contributions
 anonymously-4.903
orr-4.673
 disse-4.514
oux-4.475
church-4.356
Token
Token ID198
Feature activation+7.50e-01

Pos logit contributions
advertising+0.692
 Dresden+0.640
ary+0.627
education+0.622
defense+0.620
Neg logit contributions
 anonymously-0.661
orr-0.630
 disse-0.610
oux-0.610
church-0.592
Token(
Token ID7
Feature activation+2.36e-01

Pos logit contributions
advertising+4.852
 Dresden+4.439
education+4.355
ary+4.346
defense+4.341
Neg logit contributions
 anonymously-5.291
orr-5.005
 disse-4.861
oux-4.842
church-4.690
TokenVis
Token ID15854
Feature activation+1.66e-01

Pos logit contributions
advertising+1.585
 Dresden+1.463
education+1.429
���+1.426
ary+1.423
Neg logit contributions
 anonymously-1.635
orr-1.567
oux-1.512
 disse-1.487
church-1.464
Tokenited
Token ID863
Feature activation-4.46e-01

Pos logit contributions
advertising+0.914
 Dresden+0.834
���+0.827
ary+0.812
peria+0.809
Neg logit contributions
 anonymously-1.330
orr-1.301
oux-1.252
church-1.240
 disse-1.231
TokenY
Token ID56
Feature activation+6.67e-01

Pos logit contributions
advertising+0.143
 Dresden+0.133
���+0.132
education+0.130
defense+0.129
Neg logit contributions
 anonymously-0.147
orr-0.141
oux-0.136
 disse-0.133
church-0.132
Tokenalk
Token ID971
Feature activation+3.47e-01

Pos logit contributions
advertising+4.314
ary+3.990
peria+3.963
 Dresden+3.924
education+3.917
Neg logit contributions
 anonymously-4.639
orr-4.196
 disse-4.188
oux-4.108
church-4.054
Tokenut
Token ID315
Feature activation+2.45e-01

Pos logit contributions
advertising+2.062
 Dresden+1.914
ary+1.889
education+1.867
defense+1.860
Neg logit contributions
 anonymously-2.635
orr-2.483
oux-2.407
 disse-2.398
church-2.377
Token Tal
Token ID7193
Feature activation+1.92e-01

Pos logit contributions
advertising+1.678
 Dresden+1.583
ary+1.554
���+1.549
defense+1.537
Neg logit contributions
 anonymously-1.624
orr-1.538
oux-1.486
church-1.464
 disse-1.453
Tokenmud
Token ID41650
Feature activation+4.40e-01

Pos logit contributions
advertising+1.404
 Dresden+1.318
ary+1.295
���+1.294
defense+1.290
Neg logit contributions
 anonymously-1.201
orr-1.139
oux-1.091
 disse-1.073
church-1.058
Token Torah
Token ID33366
Feature activation+2.55e-01

Pos logit contributions
advertising+3.067
 Dresden+2.965
ary+2.929
���+2.886
defense+2.847
Neg logit contributions
 anonymously-2.788
orr-2.654
oux-2.572
church-2.510
 disse-2.481
Token,
Token ID11
Feature activation+1.23e-01

Pos logit contributions
advertising+1.769
 Dresden+1.691
���+1.659
ary+1.641
education+1.629
Neg logit contributions
 anonymously-1.673
orr-1.596
oux-1.546
church-1.503
 disse-1.493
Token
Token ID447
Feature activation-4.82e-01

Pos logit contributions
advertising+0.849
 Dresden+0.807
���+0.794
education+0.776
defense+0.774
Neg logit contributions
 anonymously-0.812
orr-0.783
oux-0.758
church-0.735
 disse-0.732
Token
Token ID251
Feature activation-4.68e-02

Pos logit contributions
 anonymously+3.556
orr+3.458
oux+3.301
church+3.260
 disse+3.212
Neg logit contributions
advertising-2.976
 Dresden-2.731
peria-2.668
education-2.660
defense-2.635
Token Genesis
Token ID18993
Feature activation+5.61e-02

Pos logit contributions
 anonymously+0.314
orr+0.301
oux+0.290
 disse+0.282
church+0.282
Neg logit contributions
advertising-0.324
 Dresden-0.303
���-0.301
education-0.294
defense-0.293
Token,
Token ID11
Feature activation+2.13e-01

Pos logit contributions
advertising+0.329
 Dresden+0.305
���+0.301
education+0.293
defense+0.292
Neg logit contributions
 anonymously-0.436
orr-0.420
oux-0.407
 disse-0.398
church-0.397
Token the
Token ID262
Feature activation-2.65e-01

Pos logit contributions
advertising+1.129
 Dresden+1.060
���+1.030
ary+1.020
defense+1.014
Neg logit contributions
 anonymously-1.379
orr-1.327
oux-1.283
church-1.259
 disse-1.256
Token point
Token ID966
Feature activation-3.14e-01

Pos logit contributions
 anonymously+1.601
orr+1.526
oux+1.458
 disse+1.421
church+1.415
Neg logit contributions
advertising-2.026
���-1.895
 Dresden-1.887
education-1.847
peria-1.844
Token where
Token ID810
Feature activation+2.44e-01

Pos logit contributions
 anonymously+2.234
orr+2.140
oux+2.045
 disse+2.009
church+2.006
Neg logit contributions
advertising-2.060
���-1.918
 Dresden-1.896
peria-1.853
education-1.848
Token this
Token ID428
Feature activation+2.75e-02

Pos logit contributions
advertising+1.629
 Dresden+1.537
ary+1.483
education+1.476
���+1.474
Neg logit contributions
 anonymously-1.663
orr-1.618
oux-1.574
church-1.527
 disse-1.503
Token sort
Token ID3297
Feature activation+3.59e-01

Pos logit contributions
advertising+0.186
 Dresden+0.174
���+0.171
education+0.168
defense+0.168
Neg logit contributions
 anonymously-0.188
orr-0.182
oux-0.176
church-0.171
 disse-0.170
Token of
Token ID286
Feature activation+1.57e-01

Pos logit contributions
advertising+2.178
 Dresden+1.967
ary+1.934
education+1.927
defense+1.915
Neg logit contributions
 anonymously-2.717
orr-2.618
oux-2.539
 disse-2.511
church-2.467
Token comparison
Token ID7208
Feature activation-1.18e-01

Pos logit contributions
advertising+1.118
 Dresden+1.054
���+1.029
education+1.016
ary+1.015
Neg logit contributions
 anonymously-1.014
orr-0.984
oux-0.955
church-0.922
 disse-0.909
Token works
Token ID2499
Feature activation+1.18e-01

Pos logit contributions
 anonymously+0.774
orr+0.742
oux+0.715
church+0.694
 disse+0.693
Neg logit contributions
advertising-0.833
 Dresden-0.778
���-0.773
education-0.756
defense-0.755
Token more
Token ID517
Feature activation-2.72e-01

Pos logit contributions
advertising+0.819
 Dresden+0.771
ary+0.751
education+0.742
defense+0.742
Neg logit contributions
 anonymously-0.772
orr-0.757
oux-0.735
church-0.710
 disse-0.696
Token as
Token ID355
Feature activation+7.19e-01

Pos logit contributions
 anonymously+1.743
orr+1.657
oux+1.581
 disse+1.549
church+1.541
Neg logit contributions
advertising-1.985
���-1.857
 Dresden-1.853
education-1.803
peria-1.798
Token a
Token ID257
Feature activation+3.23e-01

Pos logit contributions
advertising+3.625
 Dresden+3.264
ary+3.181
education+3.180
defense+3.158
Neg logit contributions
 anonymously-5.933
orr-5.897
oux-5.717
church-5.554
 disse-5.506
Token
Token ID198
Feature activation+1.06e-01

Pos logit contributions
 anonymously+0.029
orr+0.028
oux+0.027
church+0.026
 disse+0.026
Neg logit contributions
advertising-0.032
 Dresden-0.030
���-0.029
ary-0.029
education-0.029
Token
Token ID198
Feature activation+1.34e-01

Pos logit contributions
advertising+0.745
 Dresden+0.688
ary+0.677
education+0.670
defense+0.669
Neg logit contributions
 anonymously-0.715
orr-0.680
oux-0.662
 disse-0.661
church-0.641
TokenWith
Token ID3152
Feature activation+1.35e-01

Pos logit contributions
advertising+0.894
 Dresden+0.808
education+0.802
defense+0.800
ary+0.798
Neg logit contributions
 anonymously-0.945
orr-0.898
oux-0.875
 disse-0.859
church-0.840
Token the
Token ID262
Feature activation+1.36e-01

Pos logit contributions
advertising+0.904
 Dresden+0.849
���+0.833
ary+0.824
education+0.820
Neg logit contributions
 anonymously-0.931
orr-0.899
oux-0.862
church-0.843
 disse-0.836
Token console
Token ID8624
Feature activation-2.87e-01

Pos logit contributions
advertising+0.772
 Dresden+0.714
���+0.692
ary+0.688
education+0.684
Neg logit contributions
 anonymously-1.077
orr-1.048
oux-1.011
church-0.990
 disse-0.979
Token method
Token ID2446
Feature activation+3.55e-01

Pos logit contributions
 anonymously+1.407
orr+1.319
oux+1.248
 disse+1.210
church+1.192
Neg logit contributions
advertising-2.505
 Dresden-2.377
���-2.376
education-2.320
defense-2.315
Token,
Token ID11
Feature activation+1.04e-01

Pos logit contributions
advertising+2.662
 Dresden+2.512
ary+2.478
���+2.452
defense+2.442
Neg logit contributions
 anonymously-2.096
orr-2.027
oux-1.962
church-1.915
 disse-1.872
Token you
Token ID345
Feature activation-3.13e-01

Pos logit contributions
advertising+0.711
 Dresden+0.665
���+0.650
ary+0.647
education+0.644
Neg logit contributions
 anonymously-0.693
orr-0.672
oux-0.647
church-0.632
 disse-0.626
Token have
Token ID423
Feature activation+4.54e-02

Pos logit contributions
 anonymously+1.799
orr+1.712
oux+1.634
 disse+1.583
church+1.578
Neg logit contributions
advertising-2.482
 Dresden-2.334
���-2.327
education-2.276
defense-2.271
Token to
Token ID284
Feature activation+5.07e-01

Pos logit contributions
advertising+0.291
 Dresden+0.270
���+0.268
education+0.261
defense+0.261
Neg logit contributions
 anonymously-0.327
orr-0.316
oux-0.305
church-0.297
 disse-0.297
Token go
Token ID467
Feature activation+3.19e-01

Pos logit contributions
advertising+3.352
 Dresden+3.214
ary+3.204
defense+3.121
education+3.092
Neg logit contributions
 anonymously-3.178
orr-3.133
oux-3.112
church-3.039
 disse-2.921
Token had
Token ID550
Feature activation-7.95e-02

Pos logit contributions
 anonymously+1.289
orr+1.233
oux+1.185
 disse+1.149
church+1.149
Neg logit contributions
advertising-1.468
 Dresden-1.373
���-1.367
education-1.335
defense-1.333
Token known
Token ID1900
Feature activation+2.49e-02

Pos logit contributions
 anonymously+0.505
orr+0.487
oux+0.469
church+0.454
 disse+0.453
Neg logit contributions
advertising-0.576
 Dresden-0.542
���-0.536
education-0.525
defense-0.524
Token about
Token ID546
Feature activation-4.65e-01

Pos logit contributions
advertising+0.187
 Dresden+0.176
���+0.173
ary+0.171
education+0.170
Neg logit contributions
 anonymously-0.151
orr-0.146
oux-0.140
 disse-0.135
church-0.135
Token his
Token ID465
Feature activation-1.81e-01

Pos logit contributions
 anonymously+2.965
orr+2.824
oux+2.715
church+2.648
 disse+2.637
Neg logit contributions
advertising-3.346
���-3.142
 Dresden-3.114
defense-3.060
peria-3.048
Token actions
Token ID4028
Feature activation-3.52e-01

Pos logit contributions
 anonymously+1.189
orr+1.140
oux+1.097
 disse+1.065
church+1.065
Neg logit contributions
advertising-1.278
 Dresden-1.194
���-1.187
education-1.160
defense-1.158
Token and
Token ID290
Feature activation+1.77e-01

Pos logit contributions
 anonymously+2.269
orr+2.132
oux+2.048
 disse+1.991
church+1.988
Neg logit contributions
advertising-2.560
���-2.422
 Dresden-2.383
education-2.327
defense-2.324
Token had
Token ID550
Feature activation-5.89e-02

Pos logit contributions
advertising+1.257
 Dresden+1.191
���+1.158
education+1.155
ary+1.154
Neg logit contributions
 anonymously-1.107
orr-1.103
oux-1.063
church-1.016
 disse-1.002
Token taken
Token ID2077
Feature activation+1.75e-01

Pos logit contributions
 anonymously+0.373
orr+0.364
oux+0.351
church+0.338
 disse+0.337
Neg logit contributions
advertising-0.424
 Dresden-0.400
���-0.394
ary-0.387
education-0.387
Token steps
Token ID4831
Feature activation-1.93e-01

Pos logit contributions
advertising+1.235
 Dresden+1.170
���+1.139
defense+1.134
education+1.133
Neg logit contributions
 anonymously-1.121
orr-1.108
oux-1.061
church-1.025
 disse-1.024
Token to
Token ID284
Feature activation-1.91e-01

Pos logit contributions
 anonymously+1.378
orr+1.311
oux+1.264
church+1.234
 disse+1.233
Neg logit contributions
advertising-1.271
���-1.185
 Dresden-1.180
education-1.145
defense-1.140
Token get
Token ID651
Feature activation+6.06e-01

Pos logit contributions
 anonymously+1.250
orr+1.202
oux+1.159
church+1.125
 disse+1.123
Neg logit contributions
advertising-1.355
 Dresden-1.268
���-1.257
education-1.232
defense-1.230
Token
Token ID198
Feature activation-8.63e-02

Pos logit contributions
advertising+0.417
 Dresden+0.392
���+0.378
defense+0.376
ary+0.375
Neg logit contributions
 anonymously-0.430
orr-0.414
oux-0.398
 disse-0.392
church-0.390
Token
Token ID198
Feature activation+4.47e-02

Pos logit contributions
 anonymously+0.548
orr+0.524
oux+0.504
 disse+0.493
church+0.488
Neg logit contributions
advertising-0.632
 Dresden-0.590
���-0.575
education-0.574
defense-0.573
TokenDead
Token ID20489
Feature activation+4.47e-01

Pos logit contributions
advertising+0.306
 Dresden+0.282
���+0.275
defense+0.275
education+0.274
Neg logit contributions
 anonymously-0.308
orr-0.295
oux-0.284
 disse-0.279
church-0.276
Tokenline
Token ID1370
Feature activation-9.07e-02

Pos logit contributions
advertising+2.939
 Dresden+2.688
ary+2.673
defense+2.625
���+2.609
Neg logit contributions
 anonymously-3.069
orr-2.962
oux-2.909
 disse-2.796
church-2.767
Token for
Token ID329
Feature activation+4.69e-01

Pos logit contributions
 anonymously+0.537
orr+0.513
oux+0.491
church+0.476
 disse+0.476
Neg logit contributions
advertising-0.701
 Dresden-0.659
���-0.653
education-0.641
defense-0.640
Token Clubs
Token ID48901
Feature activation-1.91e-01

Pos logit contributions
advertising+3.195
 Dresden+3.047
���+2.922
defense+2.921
ary+2.886
Neg logit contributions
 anonymously-3.049
orr-3.007
oux-2.874
church-2.851
 disse-2.750
Token to
Token ID284
Feature activation+5.15e-01

Pos logit contributions
 anonymously+1.294
orr+1.237
oux+1.192
 disse+1.159
church+1.155
Neg logit contributions
advertising-1.315
���-1.226
 Dresden-1.223
education-1.193
defense-1.190
Token ask
Token ID1265
Feature activation+3.69e-01

Pos logit contributions
advertising+3.063
 Dresden+2.995
���+2.877
ary+2.858
defense+2.829
Neg logit contributions
 anonymously-3.612
orr-3.589
oux-3.494
church-3.492
 disse-3.342
Token Players
Token ID13094
Feature activation+1.24e-01

Pos logit contributions
advertising+2.491
 Dresden+2.386
defense+2.271
ary+2.271
���+2.249
Neg logit contributions
 anonymously-2.423
orr-2.408
oux-2.292
church-2.261
 disse-2.205
Token with
Token ID351
Feature activation+3.38e-01

Pos logit contributions
advertising+0.795
 Dresden+0.742
���+0.728
ary+0.714
defense+0.713
Neg logit contributions
 anonymously-0.892
orr-0.866
oux-0.834
church-0.815
 disse-0.810
Token "
Token ID366
Feature activation+3.99e-01

Pos logit contributions
advertising+2.159
 Dresden+2.026
���+1.997
ary+1.964
defense+1.947
Neg logit contributions
 anonymously-2.394
orr-2.316
oux-2.236
church-2.199
 disse-2.155
Token,
Token ID11
Feature activation+3.63e-01

Pos logit contributions
advertising+1.138
 Dresden+1.077
���+1.063
education+1.052
ary+1.044
Neg logit contributions
 anonymously-0.859
orr-0.826
oux-0.785
 disse-0.763
church-0.762
Token Ground
Token ID13706
Feature activation-1.61e-02

Pos logit contributions
advertising+2.332
 Dresden+2.268
education+2.178
���+2.154
ary+2.145
Neg logit contributions
 anonymously-2.482
orr-2.469
oux-2.341
church-2.293
 disse-2.255
Token Zero
Token ID12169
Feature activation+1.11e-01

Pos logit contributions
 anonymously+0.109
orr+0.104
oux+0.100
 disse+0.098
church+0.097
Neg logit contributions
advertising-0.111
 Dresden-0.104
���-0.103
education-0.101
defense-0.101
Token for
Token ID329
Feature activation+3.01e-01

Pos logit contributions
advertising+0.774
 Dresden+0.732
education+0.718
���+0.718
defense+0.713
Neg logit contributions
 anonymously-0.719
orr-0.702
oux-0.670
 disse-0.647
church-0.644
Token social
Token ID1919
Feature activation-7.72e-01

Pos logit contributions
advertising+2.045
 Dresden+1.953
education+1.912
���+1.903
defense+1.876
Neg logit contributions
 anonymously-1.992
orr-1.951
oux-1.873
church-1.812
 disse-1.800
Token justice
Token ID5316
Feature activation-1.39e-01

Pos logit contributions
 anonymously+4.974
orr+4.584
 disse+4.399
oux+4.344
church+4.344
Neg logit contributions
advertising-5.396
���-5.168
 Dresden-5.039
peria-5.009
ebted-4.812
Token brain
Token ID3632
Feature activation-9.64e-01

Pos logit contributions
 anonymously+0.902
orr+0.866
oux+0.833
church+0.808
 disse+0.807
Neg logit contributions
advertising-1.000
 Dresden-0.935
���-0.929
education-0.911
defense-0.908
Tokenwashing
Token ID38524
Feature activation+5.65e-02

Pos logit contributions
 anonymously+5.900
orr+5.692
 disse+5.312
oux+5.263
church+5.153
Neg logit contributions
advertising-6.861
 Dresden-6.518
���-6.487
ebted-6.232
peria-6.186
Token,
Token ID11
Feature activation-5.15e-02

Pos logit contributions
advertising+0.394
 Dresden+0.372
education+0.363
���+0.362
ary+0.359
Neg logit contributions
 anonymously-0.372
orr-0.361
oux-0.348
church-0.338
 disse-0.334
Token the
Token ID262
Feature activation+9.27e-02

Pos logit contributions
 anonymously+0.350
orr+0.339
oux+0.326
church+0.318
 disse+0.316
Neg logit contributions
advertising-0.350
 Dresden-0.328
���-0.323
education-0.319
defense-0.317
Token Social
Token ID5483
Feature activation-6.83e-01

Pos logit contributions
advertising+0.606
 Dresden+0.579
���+0.561
education+0.558
ary+0.553
Neg logit contributions
 anonymously-0.646
orr-0.630
oux-0.603
church-0.588
 disse-0.585
Token
Token ID447
Feature activation+3.13e-01

Pos logit contributions
advertising+4.291
 Dresden+4.073
���+3.912
ary+3.903
defense+3.887
Neg logit contributions
 anonymously-3.672
orr-3.491
oux-3.393
church-3.303
 disse-3.295
Token
Token ID247
Feature activation-1.25e-01

Pos logit contributions
advertising+0.944
 Dresden+0.801
���+0.751
defense+0.746
education+0.743
Neg logit contributions
 anonymously-3.304
orr-3.203
oux-3.166
 disse-3.118
church-3.092
Tokens
Token ID82
Feature activation-1.52e-02

Pos logit contributions
 anonymously+0.818
orr+0.784
oux+0.754
church+0.732
 disse+0.732
Neg logit contributions
advertising-0.892
���-0.832
 Dresden-0.832
education-0.810
defense-0.808
Token Diary
Token ID35911
Feature activation-1.79e-01

Pos logit contributions
 anonymously+0.072
orr+0.068
oux+0.064
 disse+0.062
church+0.062
Neg logit contributions
advertising-0.136
 Dresden-0.129
���-0.128
education-0.126
defense-0.125
Token.
Token ID13
Feature activation-3.08e-01

Pos logit contributions
 anonymously+1.126
orr+1.073
oux+1.031
 disse+1.000
church+0.996
Neg logit contributions
advertising-1.321
���-1.238
 Dresden-1.235
education-1.206
defense-1.203
Token
Token ID447
Feature activation-1.56e-01

Pos logit contributions
 anonymously+1.975
orr+1.897
oux+1.819
church+1.761
 disse+1.753
Neg logit contributions
advertising-2.205
���-2.084
 Dresden-2.055
education-2.010
defense-2.004
Token
Token ID251
Feature activation-5.27e-01

Pos logit contributions
 anonymously+1.341
orr+1.302
oux+1.262
church+1.237
 disse+1.232
Neg logit contributions
advertising-0.790
 Dresden-0.716
���-0.708
education-0.686
defense-0.684
Token
Token ID198
Feature activation-1.80e-01

Pos logit contributions
 anonymously+3.382
orr+3.249
oux+3.114
church+3.017
 disse+2.983
Neg logit contributions
advertising-3.733
���-3.583
 Dresden-3.457
education-3.407
defense-3.405
Token
Token ID198
Feature activation-2.93e-01

Pos logit contributions
 anonymously+1.117
orr+1.069
oux+1.026
church+0.994
 disse+0.992
Neg logit contributions
advertising-1.337
���-1.257
 Dresden-1.254
education-1.220
defense-1.218
TokenI
Token ID40
Feature activation-2.80e-01

Pos logit contributions
 anonymously+1.924
orr+1.861
oux+1.784
church+1.739
 disse+1.711
Neg logit contributions
advertising-2.031
���-1.956
 Dresden-1.923
education-1.856
defense-1.852
Token think
Token ID892
Feature activation+2.91e-02

Pos logit contributions
 anonymously+1.723
orr+1.647
oux+1.582
church+1.533
 disse+1.532
Neg logit contributions
advertising-2.103
 Dresden-1.974
���-1.961
education-1.920
defense-1.916
Token like
Token ID588
Feature activation+1.78e-01

Pos logit contributions
advertising+1.884
 Dresden+1.775
ary+1.737
���+1.713
education+1.687
Neg logit contributions
 anonymously-2.075
orr-2.026
oux-1.971
church-1.913
 disse-1.868
Token your
Token ID534
Feature activation-1.92e-01

Pos logit contributions
advertising+1.163
 Dresden+1.089
ary+1.066
���+1.056
education+1.046
Neg logit contributions
 anonymously-1.246
orr-1.206
oux-1.162
church-1.135
 disse-1.120
Token looking
Token ID2045
Feature activation-5.01e-02

Pos logit contributions
 anonymously+1.259
orr+1.207
oux+1.162
 disse+1.128
church+1.127
Neg logit contributions
advertising-1.362
 Dresden-1.273
���-1.266
education-1.236
defense-1.234
Token at
Token ID379
Feature activation-1.77e-01

Pos logit contributions
 anonymously+0.347
orr+0.335
oux+0.323
church+0.314
 disse+0.313
Neg logit contributions
advertising-0.336
 Dresden-0.313
���-0.311
education-0.303
defense-0.302
Token the
Token ID262
Feature activation-6.91e-02

Pos logit contributions
 anonymously+1.298
orr+1.249
oux+1.208
 disse+1.177
church+1.176
Neg logit contributions
advertising-1.120
 Dresden-1.037
���-1.031
education-1.004
defense-1.002
Token image
Token ID2939
Feature activation-1.24e-01

Pos logit contributions
 anonymously+0.456
orr+0.439
oux+0.423
church+0.410
 disse+0.408
Neg logit contributions
advertising-0.486
 Dresden-0.455
���-0.447
ary-0.440
education-0.440
Token through
Token ID832
Feature activation-7.87e-02

Pos logit contributions
 anonymously+0.821
orr+0.789
oux+0.760
church+0.739
 disse+0.737
Neg logit contributions
advertising-0.874
 Dresden-0.817
���-0.807
education-0.794
defense-0.792
Token a
Token ID257
Feature activation-5.57e-01

Pos logit contributions
 anonymously+0.588
orr+0.566
oux+0.548
 disse+0.534
church+0.534
Neg logit contributions
advertising-0.486
 Dresden-0.449
���-0.446
education-0.434
defense-0.433
Token mesh
Token ID19609
Feature activation-2.58e-01

Pos logit contributions
 anonymously+3.505
orr+3.357
oux+3.221
 disse+3.204
church+3.115
Neg logit contributions
advertising-4.034
���-3.809
 Dresden-3.725
education-3.689
defense-3.660
Token (
Token ID357
Feature activation-3.04e-01

Pos logit contributions
 anonymously+1.703
orr+1.632
oux+1.570
 disse+1.531
church+1.524
Neg logit contributions
advertising-1.812
 Dresden-1.694
���-1.694
education-1.647
defense-1.640
Tokenusually
Token ID23073
Feature activation-2.22e-02

Pos logit contributions
 anonymously+2.065
orr+1.999
oux+1.922
church+1.875
 disse+1.861
Neg logit contributions
advertising-2.053
���-1.939
 Dresden-1.936
education-1.860
defense-1.855
Token
Token ID447
Feature activation-8.81e-02

Pos logit contributions
advertising+0.306
 Dresden+0.285
���+0.284
education+0.277
defense+0.276
Neg logit contributions
 anonymously-0.304
orr-0.292
oux-0.281
 disse-0.273
church-0.273
Token
Token ID247
Feature activation+1.57e-01

Pos logit contributions
 anonymously+0.687
orr+0.662
oux+0.642
 disse+0.627
church+0.625
Neg logit contributions
advertising-0.518
 Dresden-0.476
���-0.473
education-0.460
defense-0.458
Tokens
Token ID82
Feature activation+2.10e-01

Pos logit contributions
advertising+1.064
 Dresden+0.989
���+0.972
education+0.968
defense+0.966
Neg logit contributions
 anonymously-1.082
orr-1.031
oux-0.995
 disse-0.974
church-0.968
Token nervous
Token ID10927
Feature activation-5.16e-02

Pos logit contributions
advertising+1.423
 Dresden+1.348
���+1.304
ary+1.301
education+1.298
Neg logit contributions
 anonymously-1.414
orr-1.363
oux-1.309
church-1.283
 disse-1.264
Tokenness
Token ID1108
Feature activation+2.47e-01

Pos logit contributions
 anonymously+0.385
orr+0.371
oux+0.358
church+0.349
 disse+0.349
Neg logit contributions
advertising-0.320
 Dresden-0.296
���-0.293
education-0.286
defense-0.286
Token about
Token ID546
Feature activation-3.24e-01

Pos logit contributions
advertising+1.847
 Dresden+1.741
ary+1.704
���+1.700
defense+1.684
Neg logit contributions
 anonymously-1.507
orr-1.447
oux-1.393
 disse-1.350
church-1.350
Token the
Token ID262
Feature activation-3.28e-01

Pos logit contributions
 anonymously+2.229
orr+2.120
oux+2.038
 disse+2.002
church+1.979
Neg logit contributions
advertising-2.201
���-2.059
 Dresden-2.027
education-1.989
peria-1.985
Token security
Token ID2324
Feature activation+4.03e-01

Pos logit contributions
 anonymously+2.249
orr+2.147
oux+2.068
 disse+2.022
church+2.011
Neg logit contributions
advertising-2.240
���-2.088
 Dresden-2.061
peria-2.016
defense-2.015
Token forces
Token ID3386
Feature activation+4.83e-02

Pos logit contributions
advertising+2.263
 Dresden+2.117
ary+2.088
education+2.042
defense+2.012
Neg logit contributions
 anonymously-3.139
orr-3.049
oux-2.970
church-2.901
 disse-2.868
Token
Token ID447
Feature activation+1.98e-01

Pos logit contributions
advertising+0.334
 Dresden+0.312
���+0.308
education+0.302
defense+0.302
Neg logit contributions
 anonymously-0.324
orr-0.312
oux-0.301
church-0.293
 disse-0.292
Token
Token ID247
Feature activation+1.39e-01

Pos logit contributions
advertising+1.011
���+0.919
 Dresden+0.919
education+0.881
defense+0.880
Neg logit contributions
 anonymously-1.698
orr-1.635
oux-1.600
 disse-1.564
church-1.561
Token,
Token ID11
Feature activation-3.92e-01

Pos logit contributions
 anonymously+1.788
orr+1.686
oux+1.617
 disse+1.567
church+1.563
Neg logit contributions
advertising-2.195
���-2.061
 Dresden-2.042
education-2.002
defense-1.994
Token Robert
Token ID5199
Feature activation+6.17e-01

Pos logit contributions
 anonymously+2.463
orr+2.321
oux+2.230
 disse+2.177
church+2.173
Neg logit contributions
advertising-2.906
���-2.724
 Dresden-2.690
education-2.637
peria-2.634
Token Gibbs
Token ID41071
Feature activation-6.63e-01

Pos logit contributions
advertising+3.743
 Dresden+3.615
ary+3.453
defense+3.381
education+3.368
Neg logit contributions
 anonymously-4.506
orr-4.285
oux-4.165
 disse-4.140
church-4.056
Token,
Token ID11
Feature activation-4.03e-01

Pos logit contributions
 anonymously+3.659
orr+3.416
oux+3.274
church+3.142
 disse+3.136
Neg logit contributions
advertising-5.374
���-4.998
 Dresden-4.982
education-4.950
peria-4.918
Token the
Token ID262
Feature activation-4.61e-01

Pos logit contributions
 anonymously+2.844
orr+2.697
oux+2.612
church+2.537
 disse+2.514
Neg logit contributions
advertising-2.662
���-2.480
 Dresden-2.437
peria-2.382
education-2.372
Token White
Token ID2635
Feature activation-1.25e+00

Pos logit contributions
 anonymously+3.063
orr+2.884
oux+2.800
 disse+2.720
church+2.720
Neg logit contributions
advertising-3.241
���-3.021
 Dresden-2.970
peria-2.920
education-2.906
Token House
Token ID2097
Feature activation-5.19e-01

Pos logit contributions
 anonymously+6.010
orr+5.707
oux+5.555
church+5.395
nect+4.924
Neg logit contributions
���-10.892
advertising-10.110
peria-9.587
amily-9.490
defense-9.444
Token press
Token ID1803
Feature activation+6.45e-01

Pos logit contributions
 anonymously+3.222
orr+3.033
oux+2.909
 disse+2.841
church+2.809
Neg logit contributions
advertising-3.801
���-3.626
 Dresden-3.586
peria-3.491
ebted-3.460
Token secretary
Token ID7705
Feature activation-4.18e-01

Pos logit contributions
advertising+2.842
ary+2.707
 Dresden+2.522
education+2.492
���+2.474
Neg logit contributions
 anonymously-5.623
orr-5.540
church-5.455
oux-5.415
 disse-5.274
Token,
Token ID11
Feature activation+2.43e-01

Pos logit contributions
 anonymously+2.188
orr+2.015
oux+1.933
 disse+1.827
church+1.811
Neg logit contributions
advertising-3.579
���-3.374
 Dresden-3.341
education-3.281
defense-3.268
Token made
Token ID925
Feature activation-2.36e-01

Pos logit contributions
advertising+1.633
 Dresden+1.553
���+1.522
ary+1.510
education+1.484
Neg logit contributions
 anonymously-1.599
orr-1.584
oux-1.528
church-1.505
 disse-1.475
Token deployed
Token ID12380
Feature activation+5.96e-01

Pos logit contributions
advertising+3.249
 Dresden+3.187
ary+3.086
���+3.035
defense+3.020
Neg logit contributions
 anonymously-3.435
orr-3.425
oux-3.301
church-3.169
 disse-3.142
Token with
Token ID351
Feature activation+2.65e-01

Pos logit contributions
advertising+3.747
 Dresden+3.575
ary+3.526
defense+3.460
education+3.391
Neg logit contributions
 anonymously-4.090
orr-4.080
oux-3.889
 disse-3.880
church-3.830
Token maximum
Token ID5415
Feature activation-2.79e-02

Pos logit contributions
advertising+1.706
 Dresden+1.637
���+1.581
ary+1.569
defense+1.556
Neg logit contributions
 anonymously-1.857
orr-1.808
oux-1.732
 disse-1.697
church-1.691
Token deployment
Token ID14833
Feature activation+1.61e-01

Pos logit contributions
 anonymously+0.199
orr+0.192
oux+0.185
 disse+0.181
church+0.180
Neg logit contributions
advertising-0.181
 Dresden-0.168
���-0.167
education-0.163
defense-0.163
Token in
Token ID287
Feature activation+1.10e-01

Pos logit contributions
advertising+1.060
 Dresden+0.995
ary+0.969
���+0.967
defense+0.962
Neg logit contributions
 anonymously-1.124
orr-1.086
oux-1.046
 disse-1.028
church-1.019
Token Roh
Token ID32694
Feature activation-1.16e+00

Pos logit contributions
advertising+0.725
 Dresden+0.691
���+0.670
defense+0.664
ary+0.662
Neg logit contributions
 anonymously-0.767
orr-0.743
oux-0.718
church-0.697
 disse-0.695
Tokent
Token ID83
Feature activation-3.11e-01

Pos logit contributions
orr+4.073
 anonymously+3.945
oux+3.806
church+3.568
ococ+3.301
Neg logit contributions
���-11.739
advertising-10.786
 Dresden-10.362
ebted-10.179
eanor-9.938
Tokenak
Token ID461
Feature activation-1.85e-01

Pos logit contributions
 anonymously+2.008
orr+1.944
oux+1.880
church+1.816
 disse+1.775
Neg logit contributions
advertising-2.212
���-2.150
 Dresden-2.079
defense-2.000
education-1.999
Token.
Token ID13
Feature activation+3.26e-02

Pos logit contributions
 anonymously+1.213
orr+1.161
oux+1.115
church+1.084
 disse+1.080
Neg logit contributions
advertising-1.320
���-1.236
 Dresden-1.228
education-1.196
defense-1.190
Token Additional
Token ID15891
Feature activation+2.93e-01

Pos logit contributions
advertising+0.213
 Dresden+0.199
���+0.194
education+0.191
defense+0.191
Neg logit contributions
 anonymously-0.233
orr-0.224
oux-0.216
 disse-0.212
church-0.211
Token columns
Token ID15180
Feature activation+6.58e-01

Pos logit contributions
advertising+1.835
 Dresden+1.741
ary+1.689
defense+1.675
education+1.664
Neg logit contributions
 anonymously-2.113
orr-2.037
oux-1.978
 disse-1.939
church-1.916
Token room
Token ID2119
Feature activation+4.30e-01

Pos logit contributions
 anonymously+1.670
orr+1.590
oux+1.525
 disse+1.478
church+1.473
Neg logit contributions
advertising-2.131
 Dresden-1.999
���-1.997
education-1.949
defense-1.945
Token internet
Token ID5230
Feature activation-2.64e-01

Pos logit contributions
advertising+3.045
 Dresden+2.899
ary+2.854
defense+2.778
���+2.760
Neg logit contributions
 anonymously-2.690
orr-2.633
oux-2.526
church-2.474
 disse-2.429
Token at
Token ID379
Feature activation-1.79e-01

Pos logit contributions
 anonymously+1.614
orr+1.534
oux+1.473
 disse+1.432
church+1.426
Neg logit contributions
advertising-1.998
 Dresden-1.873
���-1.873
education-1.825
defense-1.823
Token the
Token ID262
Feature activation-1.87e-01

Pos logit contributions
 anonymously+1.221
orr+1.174
oux+1.133
 disse+1.101
church+1.098
Neg logit contributions
advertising-1.221
���-1.131
 Dresden-1.130
education-1.103
defense-1.100
Token West
Token ID2688
Feature activation-1.30e-01

Pos logit contributions
 anonymously+1.271
orr+1.221
oux+1.177
 disse+1.146
church+1.144
Neg logit contributions
advertising-1.280
���-1.188
 Dresden-1.185
education-1.157
defense-1.153
Tokenin
Token ID259
Feature activation-1.08e+00

Pos logit contributions
 anonymously+0.870
orr+0.835
oux+0.804
church+0.783
 disse+0.780
Neg logit contributions
advertising-0.901
 Dresden-0.838
���-0.838
education-0.816
defense-0.815
Token cut
Token ID2005
Feature activation-9.50e-02

Pos logit contributions
 anonymously+7.102
orr+6.930
oux+6.753
church+6.570
 disse+6.330
Neg logit contributions
advertising-6.909
���-6.450
ebted-6.375
peria-6.295
education-6.260
Token out
Token ID503
Feature activation-3.75e-01

Pos logit contributions
 anonymously+0.623
orr+0.597
oux+0.575
 disse+0.558
church+0.558
Neg logit contributions
advertising-0.674
 Dresden-0.630
���-0.626
education-0.612
defense-0.610
Token (
Token ID357
Feature activation-1.82e-01

Pos logit contributions
 anonymously+2.391
orr+2.262
oux+2.192
 disse+2.125
church+2.112
Neg logit contributions
advertising-2.733
���-2.573
 Dresden-2.520
education-2.490
peria-2.482
Tokenas
Token ID292
Feature activation+6.11e-02

Pos logit contributions
 anonymously+1.268
orr+1.220
oux+1.177
church+1.146
 disse+1.142
Neg logit contributions
advertising-1.211
 Dresden-1.128
���-1.127
education-1.093
defense-1.090
Token it
Token ID340
Feature activation-2.22e-01

Pos logit contributions
advertising+0.410
 Dresden+0.385
���+0.378
ary+0.370
defense+0.370
Neg logit contributions
 anonymously-0.421
orr-0.406
oux-0.389
church-0.381
 disse-0.381
Token and
Token ID290
Feature activation-2.98e-01

Pos logit contributions
 anonymously+3.072
orr+2.923
oux+2.794
church+2.695
 disse+2.693
Neg logit contributions
advertising-3.796
���-3.653
 Dresden-3.551
education-3.458
defense-3.456
Token touched
Token ID12615
Feature activation-4.00e-01

Pos logit contributions
 anonymously+2.006
orr+1.916
oux+1.845
 disse+1.801
church+1.793
Neg logit contributions
advertising-2.070
���-1.926
 Dresden-1.921
education-1.872
defense-1.861
Token off
Token ID572
Feature activation+1.71e-01

Pos logit contributions
 anonymously+2.492
orr+2.364
oux+2.274
 disse+2.215
church+2.202
Neg logit contributions
advertising-2.972
���-2.796
 Dresden-2.756
education-2.708
defense-2.686
Token the
Token ID262
Feature activation+2.11e-01

Pos logit contributions
advertising+0.985
 Dresden+0.916
���+0.892
defense+0.879
ary+0.876
Neg logit contributions
 anonymously-1.341
orr-1.301
oux-1.262
church-1.230
 disse-1.220
Token political
Token ID1964
Feature activation-2.77e-02

Pos logit contributions
advertising+1.437
 Dresden+1.380
ary+1.325
���+1.325
education+1.317
Neg logit contributions
 anonymously-1.406
orr-1.351
oux-1.309
church-1.274
 disse-1.246
Token rift
Token ID36788
Feature activation-1.11e+00

Pos logit contributions
 anonymously+0.179
orr+0.172
oux+0.165
church+0.160
 disse+0.159
Neg logit contributions
advertising-0.198
 Dresden-0.186
���-0.183
education-0.181
ary-0.180
Token that
Token ID326
Feature activation+5.99e-02

Pos logit contributions
 anonymously+6.207
orr+5.864
oux+5.615
 disse+5.315
church+5.304
Neg logit contributions
advertising-8.441
���-8.264
education-7.649
 Dresden-7.631
peria-7.617
Token paraly
Token ID26901
Feature activation-2.91e-01

Pos logit contributions
advertising+0.406
 Dresden+0.387
ary+0.374
���+0.372
defense+0.370
Neg logit contributions
 anonymously-0.404
orr-0.392
oux-0.377
church-0.371
 disse-0.362
Tokenzes
Token ID12271
Feature activation-2.71e-02

Pos logit contributions
 anonymously+1.793
orr+1.721
oux+1.656
 disse+1.602
church+1.599
Neg logit contributions
advertising-2.175
���-2.039
 Dresden-2.034
education-1.977
peria-1.973
Token Palestinian
Token ID8501
Feature activation-4.77e-01

Pos logit contributions
 anonymously+0.196
orr+0.189
oux+0.184
church+0.178
 disse+0.176
Neg logit contributions
advertising-0.173
 Dresden-0.163
���-0.157
defense-0.156
ary-0.156
Token politics
Token ID4819
Feature activation+7.19e-01

Pos logit contributions
 anonymously+2.964
orr+2.837
oux+2.707
 disse+2.669
church+2.627
Neg logit contributions
advertising-3.513
���-3.332
 Dresden-3.284
peria-3.187
education-3.184
Token
Token ID198
Feature activation-1.50e-01

Pos logit contributions
 anonymously+0.886
orr+0.846
oux+0.814
 disse+0.793
church+0.789
Neg logit contributions
advertising-1.030
 Dresden-0.963
���-0.947
education-0.937
defense-0.935
TokenThe
Token ID464
Feature activation-2.38e-01

Pos logit contributions
 anonymously+1.070
orr+1.033
oux+0.994
church+0.971
 disse+0.964
Neg logit contributions
advertising-0.972
���-0.914
 Dresden-0.911
education-0.877
defense-0.875
Token statement
Token ID2643
Feature activation+1.52e-01

Pos logit contributions
 anonymously+1.579
orr+1.511
oux+1.454
 disse+1.414
church+1.413
Neg logit contributions
advertising-1.666
 Dresden-1.552
���-1.549
education-1.509
defense-1.507
Token by
Token ID416
Feature activation-4.93e-01

Pos logit contributions
advertising+1.019
 Dresden+0.966
ary+0.953
defense+0.926
���+0.924
Neg logit contributions
 anonymously-1.015
orr-1.002
oux-0.964
church-0.937
 disse-0.930
Token the
Token ID262
Feature activation-2.69e-01

Pos logit contributions
 anonymously+3.117
orr+2.959
oux+2.812
church+2.732
 disse+2.726
Neg logit contributions
advertising-3.597
���-3.376
 Dresden-3.331
peria-3.266
education-3.264
Token FDA
Token ID14672
Feature activation-1.56e+00

Pos logit contributions
 anonymously+1.860
orr+1.779
oux+1.712
 disse+1.667
church+1.665
Neg logit contributions
advertising-1.821
���-1.690
 Dresden-1.688
education-1.640
defense-1.637
Token largely
Token ID5688
Feature activation-4.81e-01

Pos logit contributions
 anonymously+9.586
 disse+8.315
orr+8.209
oux+8.105
church+7.689
Neg logit contributions
���-10.641
advertising-10.613
ebted-9.898
 Dresden-9.603
education-9.585
Token came
Token ID1625
Feature activation-2.56e-03

Pos logit contributions
 anonymously+3.179
orr+2.984
oux+2.876
 disse+2.839
church+2.793
Neg logit contributions
advertising-3.419
���-3.207
 Dresden-3.160
education-3.115
peria-3.090
Token as
Token ID355
Feature activation+1.31e-01

Pos logit contributions
 anonymously+0.019
orr+0.018
oux+0.018
 disse+0.017
church+0.017
Neg logit contributions
advertising-0.016
 Dresden-0.015
���-0.015
education-0.014
defense-0.014
Token a
Token ID257
Feature activation+2.72e-01

Pos logit contributions
advertising+0.866
 Dresden+0.807
���+0.790
education+0.780
defense+0.779
Neg logit contributions
 anonymously-0.919
orr-0.894
oux-0.865
 disse-0.837
church-0.836
Token surprise
Token ID5975
Feature activation+4.81e-01

Pos logit contributions
advertising+1.687
 Dresden+1.597
ary+1.544
defense+1.535
���+1.527
Neg logit contributions
 anonymously-1.939
orr-1.931
oux-1.868
church-1.784
 disse-1.777