topic model 0.001846703
variable model 0.001609593
model inference 0.0015920910000000001
topic models 0.0015288300000000001
generative model 0.0014486
training data 0.0014177460000000001
topic word 0.001409836
present model 0.001395313
bilistic model 0.001378874
test data 0.001354015
anchor word 0.001352161
anchor words 0.001349902
purpose language 0.001293391
other words 0.001286757
words algorithm 0.001244988
anchor topic 0.001244771
topic anchor 0.001244771
new words 0.001214285
model 0.00119548
anchor anchor 0.001187096
probabilistic models 0.0011861839999999998
other topic 0.001181626
topic matrix 0.001171503
variables models 0.001131073
test time 0.001122724
single word 0.001119429
subsequent models 0.001105357
regularized models 0.001094931
language 0.00109219
anchor algorithm 0.0010821820000000001
topic modeling 0.00107836
topic distribution 0.001077108
robust models 0.001072291
data reconstruction 0.001070885
models seeds 0.001061068
interpretable models 0.001061068
data likelihood 0.001052782
development data 0.001051365
common words 0.001049973
similar words 0.001044142
anchor methods 0.001041247
words original 0.001029115
different regulariza 0.001023925
specific word 0.001018675
unseen data 0.001014375
opment data 0.001013283
rethinking data 0.001012557
approach anchor 0.0010121689999999998
word probabilities 9.95713E-4
topic number 9.95348E-4
test set 9.93367E-4
word algorithms 9.76887E-4
relevant words 9.73543E-4
word game 9.72448E-4
word count 9.70394E-4
frequent words 9.69109E-4
word distribu 9.6634E-4
bold words 9.64476E-4
liwc word 9.64025E-4
conditional word 9.63953E-4
top words 9.58048E-4
fixed word 9.554839999999999E-4
word indexes 9.55004E-4
polysemous words 9.5437E-4
word vec 9.50622E-4
chor word 9.47303E-4
probable words 9.46833E-4
word candidate 9.45636E-4
chor words 9.45044E-4
word wsk 9.432900000000001E-4
shared words 9.41526E-4
colored words 9.40614E-4
anchor method 9.191949999999999E-4
modeling algorithm 9.157710000000001E-4
new methods 9.0563E-4
topic mod 9.026990000000001E-4
topic comparison 8.971280000000001E-4
good anchor 8.878849999999999E-4
number figure 8.84836E-4
topic interpretability 8.788400000000001E-4
models 8.77607E-4
topic coefficients 8.71984E-4
topic coherence 8.71056E-4
topic distributions 8.709330000000001E-4
anchor coefficient 8.70025E-4
original anchor 8.66309E-4
human users 8.630220000000001E-4
new regularization 8.49163E-4
unregularized topic 8.41096E-4
old topic 8.369670000000001E-4
topic coeffi 8.36578E-4
topic shared 8.36395E-4
modern topic 8.363330000000001E-4
topic representatives 8.34467E-4
topic interpretabil 8.34467E-4
corpus train 8.336299999999999E-4
beta anchor 8.323499999999999E-4
training reconstruction 8.30033E-4
time people 8.286890000000001E-4
new research 8.202870000000001E-4
