This simulation was constructed in the same manner as the bulk simulations detailed on pages 603-604 of Fusi et al. (2005). [1] However, the probability that a plasticity event would be potentiating rather than depressing (f_plus, in the paper) was increased from 0.5 to 0.66, 0.75, and 0.95 from t = 20 to t = 50 (which means that f_minus was decreased from 0.5 to 0.34, 0.25, and 0.05). At all other points, the f_plus and f_minus parameters were kept constant at 0.5.

This simulation was constructed in the same manner as the bulk simulations detailed on pages 603-604 of Fusi et al. (2005). [1] However, the number of cascade substates in the weak synaptic state (brown circles, in the paper figure) ranged from 5, 10, and 20 states. The number of cascade substates in the strong synaptic state (teal circles, in the paper figure) was held constant at 10 states. The transition probability parameters q were separated between the strong and weak synaptic states due to this asymmetry, but the probabilities were still calculated using the formulas provided in the paper.

Please feel free to email me at argha (dot) bandyopadhyay (at) gmail (dot) com if you would like me to send you the Python scripts that were used for this analysis!