Skip to main content
. 2020 Aug 31;31(1):233–247. doi: 10.1093/cercor/bhaa222

Figure 3.

Figure 3

Reducing entropy in prediction before (left panel) and after (right panel) a verb is incorporated into the SNP context. The topic distributions on the top are the semantic blends of predicted CNs by SNP and SNP + verb, respectively. Entropy associated with each of the two distributions is also described. The word boxes below the distributions show a set of preferred words based on the predicted topics.