vector model 0.003077383
first model 0.002600321
model weights 0.0025233210000000002
word representations 0.002462482
such word 0.002458586
model parameters 0.002437322
model implementation 0.0024025920000000003
model perfor 0.00239909
doc model 0.0023933310000000003
semantic representations 0.002389022
model variables 0.0023841500000000002
all model 0.002377785
add model 0.0023770650000000003
semantic representation 0.002265601
model 0.00216299
semantic similarity 0.00215254
training data 0.002142953
semantic information 0.0021245450000000003
word vectors 0.002124087
semantic space 0.002079007
word figure 0.002072028
semantic transfer 0.001945071
word embeddings 0.001942754
word level 0.001853945
word alignment 0.0018465110000000001
word baselines 0.001834804
test data 0.001834069
word alignments 0.001802615
semantic rep 0.001792963
semantic representa 0.0017638760000000002
semantic con 0.001762493
semantic signal 0.001746603
semantic rela 0.001735061
such models 0.001732482
semantic signals 0.001731671
semantic grounding 0.001729009
second language 0.0017070359999999999
parallel data 0.0016882120000000001
language pairs 0.001634704
additional data 0.001584282
single language 0.0015612
language pair 0.0015584330000000001
multilingual data 0.001544184
development data 0.00153319
paired language 0.001526014
training corpus 0.0015158860000000001
language processing 0.001515881
natural language 0.0015131020000000001
gual data 0.001502548
german data 0.001496237
language acquisition 0.001476091
vector composition 0.001472847
annotated data 0.001472658
language modelling 0.00146356
lel data 0.001456035
europarl data 0.001455866
tilingual data 0.001450104
opment data 0.001450104
representation learning 0.001438618
composition models 0.00142224
compositional vector 0.0014146879999999999
compositional representations 0.0013728870000000002
parallel training 0.001354985
learning method 0.001350618
similar words 0.001327974
such approaches 0.0013211949999999998
sentence vectors 0.001295085
individual words 0.001276857
syntactic information 0.001270564
compositional sentence 0.001261183
new training 0.001248097
classification task 0.001247707
language 0.00124661
tional vector 0.001239211
individual models 0.001210614
such methods 0.001208811
learning function 0.001207472
test corpus 0.001207002
classifier training 0.001197758
document classification 0.001189263
mode vector 0.001176163
supervised training 0.001175269
sentence pairs 0.001148982
such rep 0.001145229
mantic representations 0.001140777
classification tasks 0.0011384770000000002
level representations 0.0011366470000000002
tic representations 0.001135165
training signal 0.001135036
document composition 0.001130675
continuous vector 0.001130226
tor representations 0.001125601
positional representations 0.001123069
training sizes 0.001118717
training epoch 0.001116831
classification results 0.0011165160000000001
lingual representations 0.001112149
additional sentence 0.00110708
parallel sentences 0.001102862
doc models 0.0010941269999999999
