t2
чето я запутался
в заключении они громят модели, но в abstract написано обратное :
In this paper, we performsuch an extensive evaluation, on a widerange of lexical semantics tasks and acrossmany parameter settings. The results, toour own surprise, show that the buzz isfully justified, as the context-predictingmodels obtain a thorough and resoundingvictory against their count-based counter-parts.
