What is and what is not a theory of context correlations

Recently a feed-forward model was put forward to account for the phenomenon of context correlations (observed by Miyashita and by Yakovlev et al) to replace the attractor picture developed for these findings. It is argued that the new proposal misses the main issues posed by the experimental situation as well as the salient aspects of the attractor scenario, which combines delay activity neural dynamics with learning at-a-distance.