NEAT is an AI model developed by UT Austin (United States), first published in June 2002. It works in the other domain, on tasks such as pole balancing.
Epoch AI has no training-compute estimate for this model.
The reference paper has 3,608 citations.