An ABSURDIST model vindicates a venerable theory.
Can relations between concepts in one system be sufficient on their own for locating matching concepts in another system? This question in concept semantics has been contentious. Now, contrary to one important theory, the answer appears to be yes. Robert Goldstone's and Brian Rogosky's new neural-net program, ABSURDIST, has demonstrated what purely theoretical considerations suggested could not be accomplished, showing once again the power of experimental computer modeling.
Journal Article↗