state function 0.0026277699999999998
function learning 0.002396248
loss function 0.002184422
different function 0.0020970249999999998
same function 0.002081975
function space 0.002013611
ing function 0.002005061
pruning function 0.0019591269999999997
cost function 0.0018786649999999998
level function 0.001869143
heuristic function 0.0018513969999999999
approximate function 0.001829367
ranking function 0.001823757
coreference system 0.00181883
function fscore 0.001804532
function fprune 0.001782463
scoring function 0.001761248
function fˆprune 0.0017512369999999999
coring function 0.0017512369999999999
function fˆscore 0.0017512369999999999
other coreference 0.0017385130000000001
function 0.00153963
greedy coreference 0.00151642
next state 0.001473477
current state 0.001468261
system mention 0.001464536
same search 0.001463087
greedy search 0.0014289419999999999
search space 0.001394723
coreference output 0.001392794
example state 0.001355882
coreference clusters 0.001355771
learning problem 0.0013466469999999999
terminal state 0.001344757
initial state 0.0013440330000000001
previous coreference 0.001343515
tial state 0.001339188
state sˆi 0.001306163
minal state 0.001300219
coreference resolution 0.0012938210000000001
training approach 0.001287666
other actions 0.001270396
setup coreference 0.001261847
coreference decisions 0.00125618
learning algorithms 0.0012458489999999998
other word 0.0012454010000000001
return coreference 0.001232412
clustering approach 0.001231839
incremental coreference 0.001230023
bad coreference 0.001227608
coreference resolu 0.001225582
coreference scorer 0.0012244579999999999
coreference resolver 0.001224219
score approach 0.001216241
search step 0.001191944
training set 0.001183344
search architecture 0.0011833030000000001
feature functions 0.0011823670000000001
possible actions 0.0011600550000000001
learning prob 0.001151753
first system 0.0011495400000000001
domain constraints 0.001146477
rank learning 0.001145309
filter learning 0.001144962
search trajectory 0.001137819
task loss 0.001132901
learning problems 0.001131805
learning procedure 0.001129915
training data 0.001118763
loss score 0.001110894
ing loss 0.001110223
imitation learning 0.001109287
loss functions 0.001103994
new action 0.001103594
current feature 0.001103286
learning rate 0.001099056
test set 0.001089864
state 0.00108814
imbalanced learning 0.0010688989999999999
realizable learning 0.0010688989999999999
pruning loss 0.001064289
novel approach 0.001051553
other components 0.001050811
level information 0.001048112
extra system 0.001041896
next mention 0.001039263
feature representation 0.001032533
classifier approach 0.001029221
reconcile system 0.001025675
extract system 0.001023357
feature vector 0.001022551
additional information 0.001011193
coreference 0.00100822
small loss 0.00100817
approach our 0.0010068540000000002
mention pair 9.99823E-4
new actions 9.96903E-4
indicator feature 9.95815E-4
large set 9.88273E-4
same features 9.875729999999998E-4
