model.arg2 = spa.State(D, vocab=vocab_concepts,
feedback=0) #argument 2 from input
model.answer = spa.State(
D, vocab=vocab_concepts, feedback=1, feedback_synapse=.05
) #result from retrieval (or counting in the full model)
#set the inputs to the model (bypassing the need for a visual system)
model.input = spa.Input(goal=goal_func,
target=target_func,
arg1=arg1_func,
arg2=arg2_func)
#Declarative Memory
model.declarative = assoc_mem_acc.AssociativeMemoryAccumulator(
input_vocab=vocab_problems,
wta_output=True,
wta_inhibit_scale=10,
threshold=.5)
#Comparison
model.comparison = compare_acc_zbrodoff.CompareAccumulator(
vocab_compare=vocab_concepts, status_scale=.6, status_feedback=.6)
#Motor
model.motor = spa.State(dimensions=Dlow, vocab=vocab_motor)
#Basal Ganglia & Thalamus
actions = spa.Actions(
#encode and retrieve
a_retrieve=
#Imaginal: a network with three slots: arg1, arg2, and answer.
model.imaginal = nengo.Network(seed=fseed)
with model.imaginal:
model.arg1 = spa.State(D, vocab=vocab_concepts,feedback=0) #argument 1 from input /for count feedback=1
model.arg2 = spa.State(D, vocab=vocab_concepts,feedback=0) #argument 2 from input /for count feedback=1
model.answer = spa.State(D, vocab=vocab_concepts,feedback=1,feedback_synapse=.05) # /for count feedback=.8
#model.answer = spa.State(D, vocab=vocab_concepts,feedback=1,feedback_synapse=.05) # /for count feedback=.8
#set the inputs to the model (bypassing the need for a visual system)
#model.input = spa.Input(goal=goal_func, target=target_func, arg1=arg1_func, arg2=arg2_func)
#from count model
model.input = spa.Input(vision=vision_input_func,goal=goal_input_func)
#Number memory
model.number_memory = assoc_mem_acc.AssociativeMemoryAccumulator(input_vocab = vocab_numbers, wta_output=True, status_scale=.7,threshold=.1,status_feedback=.3)
#Alphabet memory
model.letter_memory = assoc_mem_acc.AssociativeMemoryAccumulator(input_vocab = vocab_letters, wta_output=True, status_scale=.7,threshold=.2,status_feedback=.3)
#Comparison
model.comparison = compare_acc_zbrodoff.CompareAccumulator(vocab_compare = vocab_concepts,status_scale = .4,status_feedback = .2, status_feedback_synapse=.05, threshold_cleanup=.1)
#final Comparison
model.comparison2 = compare_acc_zbrodoff.CompareAccumulator(vocab_compare = vocab_concepts,status_scale = .6,status_feedback = .2, status_feedback_synapse=.05, threshold_cleanup=.1)
#Motor
model.motor = spa.State(Dlow,vocab=vocab_motor,feedback=1) #motor state
#bcm_model
model.arg2 = spa.State(
D, vocab=vocab_concepts,
feedback=0) #argument 2 from input /for count feedback=1
model.answer = spa.State(
D, vocab=vocab_concepts, feedback=1,
feedback_synapse=.05) # /for count feedback=.8
#set the inputs to the model (bypassing the need for a visual system)
#model.input = spa.Input(goal=goal_func, target=target_func, arg1=arg1_func, arg2=arg2_func)
#from count model
model.input = spa.Input(vision=vision_input_func, goal=goal_input_func)
#Number memory
model.number_memory = assoc_mem_acc.AssociativeMemoryAccumulator(
input_vocab=vocab_numbers,
wta_output=True,
status_scale=.7,
threshold=.1,
status_feedback=.3)
#nengo.Connection(model.number_memory.output,model.number_memory.input,transform=.2) #feedback on number memory
#Alphabet memory
model.letter_memory = assoc_mem_acc.AssociativeMemoryAccumulator(
input_vocab=vocab_letters,
wta_output=True,
status_scale=.7,
threshold=.2,
status_feedback=.3)
#nengo.Connection(model.letter_memory.output,model.letter_memory.input,transform=.2) #feedback on letter memory
#Declarative Memory
model.declarative = assoc_mem_acc.AssociativeMemoryAccumulator(