Brainscape's Knowledge GenomeTM

Browse over 1 million classes created by top students, professors, publishers, and experts.


See full index

hybw > L8 computational neusci > Flashcards

L8 computational neusci Flashcards

1
Q

Does all scientific activity revolves around the creation, testing, refinement and debate of scientific models?

A

Yes

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

What is Newtonian Mechanics?

A

Able to use a small number of very simple mathematical equations to describe the movement of planets with remarkable accuracy

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

What is parsimony?

A

Decide what is important and ignore the rest

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

What is explanatory?

A

Give good descriptions

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

What is predictive power and falsifiability

A

Allow precise predictions

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

Models of the brain have to strike the right balance between ______/______ and ______ of the explanations or predictions they make

A

simplicity; parsimony; accuracy

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

What do compartmental models do?

A

stimulate individual neurons as lots (sometimes thousands) of connected tubelets and each has its equivalent circuit

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q 
Purkinje cells:
Number of compartments
Number of ion channels
Number of different voltage-dependent channels distributed across
Number of synapses
What does it do?
A

4550;8021; 10; 3500; reproduces electrical current injection and synaptic responses

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

Prof Markram‘s Blue Brain project tries to model parts of the neocortex by running vast _________ on massive super computers

A

compartmental models

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

What did the Blue Brain project ignore?

A

make the model parsimonious by leaving out as much irrelevant detail as possible

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

Why there is no method for trying to decide what details are relevant and which are not for Blue Brain project?

A

It is unclear what aspects of brain function the blue brain is suppose to model

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

Did Blue Brain yield great return?

A

No

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

Which of the following are spiking models/ non-spiking models (rate based)?:
Integrate and Fire
Linear threshold or linear-non linear neurons
linear neurons
the Izhikevich neurons

A

spiking model:
Integrate and Fire
the Izhikevich neurons

non-spiking models (rate based):
Linear threshold or linear-non linear neurons
linear neurons

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

It is useful to think of neurons as ______

A

in a first order approximation

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

How to compute neurons?

A

take all synaptic inputs (many of which may be zero most of the time) -> multiply each input by its ‘synaptic weight’ -> sum -> use a threshold criterion/ a linear/ sigmoidal function to decide how much output to send given the weighted sum of inputs

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

What is the major property of biological neurons?

A

firing action potentials in response to synaptic input

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
17
Q

Essential features in simplified models of spiking neurons.

A

Action potential’s shape is same each time, so just treat the occurence of an AP as a binary 1
All EPSPs and IPSPs eventually get summed in the cell body, so no need to model dendrites

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
18
Q

Integrate-and-fire model 3 properties.

A

Point neuron; resistance+capacitance; synaptic input

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
19
Q

Which is simpler? Rate-based/ spiking neurons?

A

Rate-based

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
20
Q

Rate-based neuron models’ inputs are ______ & ______ to calculate the ‘activation’

A

weighted; summed

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
21
Q

Output in rate-based models could be _____/ _____/ ______/ ________

A

sigmoid; threshold; threshold linear function; linear function

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
22
Q

Computing units used in artificial neural network can be regarded as models of _________

A

biological neurons

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
23
Q

Activation function captures essence of a neuron’s spiking response using ____ & _____

A

Binary (0/1): neuron is silent/ firing - McCulloch-Pitts threshold logic unit
Analog: number represents neuron firing rate -piecewise linear/ sigmoid

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
24
Q

McCullough-Pitts Neurons is also known as.

A

Binary Threshold Neurons

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
25
Q

What is the simplest neuron model?

A

Binary threshold neuron

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
26
Q

What do binary threshold neuron model receive?

A

Inputs form synapses that have positive/ negative weights (excitatory/ inhibitory synapses)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
27
Q

What is the activation of binary threshold neuron model?

A

Weighted sum of inputs

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
28
Q

How does binary threshold neuron produce a binary output of 0/1?

A

Depends on whether the activation is greater than a given threshold

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
29
Q

How is binary threshold neuron similar to biological neurons?

A

Summate synaptic currents in their dendrites and soma, and fire an action potential or not depending on whether there is depolarization

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
30
Q

Binary threshold neurons are “_______”

A

Logic Gates

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
31
Q

How to make it possible to make networks compute any function?

A

By setting synaptic weights on networks of binary threshold neurons appropriately

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
32
Q

Linear and linear-nonlinear neurons weight and sum inputs the same way like _______ do to compute activation

A

binary threshold neurons

33
Q

Linear and linear-nonlinear produce a _______ output (their firing rate) which is either just ______ to the activation (____) or is a ______, often ____, function of the activation

A

real valued; proportional; linear; non-linear; sigmoidal

34
Q

Name 3 functions of the sigmoids

A

Logistic function
Arctan function
Auditory nerve fiber rate-level functions

35
Q

Which model can model the input-output relationships of sensory neurons as ‘weight matrix’?

A

Receptive field model

36
Q

What is ‘weight matrix’

A

It says what the synaptic weight of each given ‘spatial’ or ‘auditory frequency channel’ would be

37
Q

Name 3 examples of receptive field models.

A

Center-surround receptive fields in retina or LGN
Simple cell receptive fields in V1
Spectro-temporal receptive fields in auditory structures

38
Q

What functions do the linear or LN models serve?

A

Predictive power and falsifiability

39
Q

What do the linear or LN models make easy to do?

A

Easy to predict what the response to any arbitrary stimulus should be

40
Q

What do the Difference of Gaussians (DoG) model of retinal ganglion cell receptive fields posit?

A

The weight of a spot of light on the retina as a function of distance from the receptive center is given by a broad inhibitory (negative, surround) bell curve and a sharp excitatory (positive, center) bell curve

41
Q

The centre-surround structure of retinal ganglion cells turns them into “______”

A

spatial frequency filters

42
Q

Larger RGC receptive fields are tuned to “________” structure in the visual scene, while smaller RFs are tuned to ______ structure

A

coarsely grained; fine-grained

43
Q

What kind of filter does the Gabor Filter Model of V1 simple cells?

A

Linear filter

44
Q

Binaural STRFs in A1 can predict ______ & _______

A

spatial tuning; how the neurons respond, as a function of time, to the sounds coming from right/left ear

45
Q

Binaural STRFs in A1 is a hugely ______ model. Because _____

A

parsimonious; it takes all the complicated auditory brain stem away for simplification

46
Q

Rosenblatt’s Perceptron is the first “______”, basically just a _________

A

artificial neural net; single binary threshold neuron

47
Q

Rosenblatt’s Perceptron is used as a “______”; (output goes high if inouts belong to class ___, low for class ______)

A

classifier; 1; 2

48
Q

Rosenblatt’s Perceptron’s weight are set through ____ by thousands of examples: If the desired output for a given input set is high, then increase the weight of those inputs that were ___, but decrease those were ___

A

training; high; low

49
Q

Rosenblatt’s Perceptron is a form of _____ learning

A

associative

50
Q

A ______ perceptron cannnot learn functions like ‘exclusive or’ but ______ perceptrons can in principle compute any arbitrary function

A

single layer; multilayer

51
Q

Rosenblatt’s Perceptron uses a ____ function to compute binary outputs. Multilayer perceptrons use _____ (____/____)

A

threshold; sigmoids; arctan; logistic

52
Q

What does multilayer perceptrons mean?

A

Have one or more ‘hidden layers’ of artificial neurons between input and output

53
Q

Weights are still set through ______ and ______, but errors during training stages must be ‘____’ to adjust the weights of the hidden layer units

A

training; associative learning; backpropagated

54
Q

Are there any input-output mapping that a sufficiently big multilayer perceptron cannot learn?

A

No

55
Q

What is deep neural network?

A

multilayer perceptrons with more than one hidden layer

56
Q

After training, neurons in subsequent layers may become sensitive to ‘features’ of ever greater _______ and ______

A

complexity; abstraction

57
Q

In ThreeBlueOneBrown: first hidden layer: ______

second hidden layer: _______

A

sensitive to little edges; sensitive to more complex shapes

58
Q

In both _____neural nets, the features the hidden neurons are actually sensitive to, may be obscure.

A

real and artificial

59
Q

Deep neural networks for image classification often have a hierarchy of layers sensitive to _________.

A

high level features

60
Q

Deep neural networks for image classification is equivalent to various parts of the _______ and the_________ or _________.

A

ventral ‘what’ stream; dorsal ‘where’ stream; color areas of extrastriate visual areas

61
Q

A _ layer RNN with only ____ hidden neurons trained on 4.4MB text file containing all the collected works of Shakespeare produced artificial Shakespeare.

A

3; 512

62
Q

Multilayer perceptrons need to be trained on very ___ datasets, comprising tens of thousands to millions of ______, and for each example the ______, ‘_____’ answer must be

A

large; examples; desired; correct

63
Q

Multilayer perceptrons: The synaptic weights of the network are adjusted ______ to make the network output more similar to the _______________

A

gradually; correct answers in the training set

64
Q

Multilayer perceptrons: “Errors” (differences between current and correct output) must “________” through the ___________

A

backpropagate; layers of the network

65
Q

What is the criticism of multilayer perceptrons?

A

Real brains do not learn like this, they don’t backpropagate/ teachers give us thousands of examples

66
Q

What is unsupervised learning?

A

Learn types of things without ‘teacher’ input which tells them ‘correct’ answers

67
Q

What will network under unsupervised learning discover?

A

natural ‘clustering’ of certain features

68
Q

What way of learning is more biological?

A

unsupervised learning

69
Q

What is reinforcement learning?

A

a network is not told what the output for a certain input should be, but it is simply ‘rewarded’ for ‘good’ outputs or ‘punished’ for ‘bad’ outputs

70
Q

Reinforcement learning will try to predict which output (behaviour) lead to ________.

A

maximum reward and minimum punishment

71
Q

What in the brain is thought to implement reinforcement learning?

A

Levels of the neurotransmitter dopamine representing ‘reward prediction error’

72
Q

Temporal difference gammon is a very modestly sized network of less than ____ artificial neurons

A

300

73
Q

Temporal difference gammon can beat more human players at _______

A

backgammon

74
Q

Neural network, whether biological/ artificial, compute by _______________

A

combining weighted inputs to calculate outputs

75
Q

Networks must often be trained by _______

A

adjusting the weights to produce useful outputs

76
Q

The trained weight matrix is a form of “________”

A

long term memory

77
Q

Recurrent neural networks also have “ ________” through activity running _______

A

short term memory; in loops

78
Q

Subsequent layers of the network tend to become sensitive to _______________ of the inputs

A

increasingly higher order features

hybw (12 decks)

Brainscape helps you reach your goals faster, through stronger study habits.
© 2024 Bold Learning Solutions. Terms and Conditions