There was a similar experiment where a research ground up flatworms and fed them to other flatworms and they learned the behavior faster. [1]
I suspect the worms aren't "learning" anything, but are simply adapting an inbuilt behavior based upon local conditions. If local conditions cause a hormone or epigentic shift, that would appear the behavior was learned and could be preserved when the brain was regrown. It would also explain how feeding the remains of other worms could "teach" it something. In reality it could just be an adjustment of hormone levels.