| T87-1014 | useful network is one that does an | identity mapping | ( Figure 4 ) . The network has |
| D15-1002 | model has to learn " only " an | identity mapping | . This is not trivial , though |
| T87-1014 | means one 's mood is devated ? Our | identity mapping | networks can be put to good use |
| W10-2501 | id is the suitable ( partial ) | identity mapping | . It follows that s ∈ / |
| P06-1122 | were : • standard : the | identity mapping | ; • max-pref : a prefix |
| N04-1008 | to assign high probabilities to | identity mappings | ( and therefore be better distributed |
| W04-3219 | languages are identical , we added an | identity mapping | for each source word si : an |
| P15-1006 | are able to recover many simple | identity mappings | without using lexical similarity |
| P11-1002 | linguistic knowledge . We use | identity mappings | for numeric values ( for ex - |
| P00-1018 | to assume that negation is an | identity mapping | . Both potential changes preserve |
| N10-1135 | to the original ones than the | identity mapping | . For example , on our data , |
| P06-2036 | KARIZE ( 7r ) initialize with | identity mapping | h ← hmin ← hmax ← |
| W00-0405 | networks of related lexical items ( | identity mappings | , synonyms , hypernyms , etc. |
| W08-0703 | form given underlying encourages | identity mapping | , / x / → -LSB- x -RSB- |
