Loading...

Leabra stands for local, error-driven and associative, biologically-realistic algorithm, and was originally developed in O’Reilly (1996). It is the basis for the Computational Cognitive Neuroscience textbook (O’Reilly et al., 2012) and its predecessor (O’Reilly & Munakata, 2000). Much of the content here is now based on Axon which uses discrete spiking instead of the rate-code activations used in Leabra. Otherwise, all of the same principles and motivations behind Leabra apply to Axon, which is the next logical progression of the overall effort to develop a computationally tractable and biologically accurate model of the mammalian brain.

• • Book
• • Sims
• • Teaching