Week 2: deep learning in biological neurons and networks Flashcards
What does the cell membrane of a neuron do?
The cell membrane of neurons are specialised for electrical computations
What are ions?
Atoms with electrical charges
How do ions pass through the cell membrane?
Call membranes do not allow ions through. Ions must move through the ion channel or a pump
How is an imbalance of ions across the cell membrane achieved?
Sodium/potassium pumps establish an imbalance of ions across the membrane
What is the concentration of sodium and potassium around the neuron cell?
- much more sodium (Na+) on the outside of the cell
- a little more potassium (K+) inside the sell
How do ions move towards balance?
Opening ion channels allows ions to pass though towards balance. There are different ions for sodium and potassium ions
What is the resting state of the cell membrane?
Voltage of -70mV
Is the inside or outside of the cell more positively/negatively charged?
The inside of the cell is negatively charged compared to the outside
What do neurotransmitters do to a post-synaptic neuron?
Neurotransmitters released from a pre-synaptic cell can excite or inhibit activity in the post-synaptic neuron
- glutamate excites activity by opening the ion channel, causing the cell membrane to depolarise
- gamma-amino butyrate acid (GABA) inhibits activity by causing chloride to enter the cell, making the cell hyperpolarised
- when a neurotransmitter binds, the ligand-gated ion channels open
What are two ways that an ion channel can open?
- When a neurotransmitter binds to the ion channel, the ion channel changes (ligand-gated channels) shape and opens
- The ion channel opens due to a change in membrane voltage (voltage-gated channels)
What causes voltage-gated ion channels to open?
If the membrane polarisation reaches a threshold (due to many excitatory inputs), the voltage-gated ion channels will open
What happens when the cell membrane reaches a certain threshold?
Causes an extreme depolarisation of the cell and the voltage-gated ion channels will open. This causes the action potential to “fire” meaning the cell passes on an electrical signal in the form of neurotransmitters to other cells. Passes down the axon to the next layer of neurons
What is the strength of the firing rate of the neuron?
The firing rate increases as the strength of the inputs increases, at some point there is a maximum firing rate
What is the threshold in biological neurons?
Around -15mV
What happens to the cells once the action potential membrane has passed?
The ion channels are closed by their outer segment
How are neurotransmitters released?
The arriving action potential causes voltage-gated ion channels to open, calcium ions flow into the cell and cause proteins in the axon terminal to change shape, releasing the neurotransmitter
What is Hebb’s postulate and what does it mean?
“cells that fire together, wire together”
-when neurons fire together over and over, the brain strengthens responses to these patterns and learns to trigger these neurons together. humans learn by an unsupervised process by recognising patterns of activity we have seen before
What is long term potentiation?
The lasting enhancement of synaptic connection by co-activation of presynaptic and postsynaptic neurons
What do the network layers of an artificial network represent?
Layers of neurons at different levels of synapses
What does the filter of an artificial network represent?
The integration of excitatory post-synaptic potentials and inhibitory post-synaptic potentials across the dendrite tree
What does the threshold of an artificial network represent?
The activation of voltage-gated sodium channels by above-threshold depolarisations by summation of EPSPs and IPSPs
What four sensors does the eye have for light? where are they all located?
Three types of colour-sensitive cones sensors (red, green, blue light intensity) and one type of rod sensor (overall light intensity), called photoreceptors. In central vision we find only cones, further out mostly rods. density drops off with distance from central vision
What is a cortical map representation?
Where the visual image is “mapped” onto the cortical surface
What is important when the eyes map an image onto the cortex?
Neighbouring locations must be represented next to each other so that the spatial extent of a filter represents a continuous piece of image
Why is it important that spatial relationships are maintained when mapping an image onto the cortex?
- allows analysis of spatially-restricted patterns
- the dendrites of the next layer will sample from this layer and they have a limited extent
How does the brain sample an image?
It magnifies the representation of important parts of the space: central vision, hands and face
Why does the brain magnify some parts of an image?
To balance detailed analysis with computational load
-and because we can change what we sample in detail by moving our hands and eyes
How does an artificial network sample an image differently to the brain?
Artificial network processes the whole image in detail, a human decides where to point the camera
What is the receptive field in the biological system?
The part of the input image (the spatial extent of the image) that produces a positive response in a biological neuron
What is the suppressive surround?
The part of the image that suppresses a response in a biological neuron
How is an image first transformed in the brain?
The brain converts the amount of light at each location to the amount of contrast. An all white image would produce zero response but a bright point in the middle would have a positive response in the middle surrounded by a negative response`
What does the receptive field refer to in an artificial network?
The spatial extent of a convolutional filter in the input image or previous feature map. this is a programmed parameter
What is a limitation of knowledge of the biological receptive field?
It is hard to determine the spatial spread of a cells input.
How does the activation of cones tell us about the image’s colour?
The relative activation of different cones must be compared to determine the colour
What do retinal ganglion cells do?
Acts as filters, breaks down the image into different components at different spatial frequencies, different spatial scales, redness, greenness, blueness, yellowness, lightness, darkness and many combinations of these attributes
What are the properties of bistratified retinal ganglion cells?
No suppressive surround around the activating centre, but limited spatial spread
What are the properties of midget ganglion cells?
Small receptive fields and small suppressive surrounds with different colour inputs to the receptive field and the surround
How are retinal ganglion cells as filters different from artificial network filters?
As we go from central to peripheral vision, the spatial extent of these filters increases, so the filters are not the same over the whole visual field. there can also be multiple receptive field sizes at the same location
Why are there different sized receptive fields in the biological system?
The retina can break down any image into different components at different spatial frequencies and spatial scales, and can use them to see different things but put them back together to recreate the original image
What is normalisation in the biological system?
The inhibition of activity by average nearby activity
What is pooling in the biological system?
There is no distinct problem of computational overload in the biological system. progressive layers have smaller representations and more feature maps, so fewer neurons are used to sample the previous layer
How can filters sometimes end up with similar weights in the biological system?
If a filter structure develops at one part of the feature map to compute something useful, similar filter structures are likely to develop elsewhere
How do weights develop in the biological system?
Weights at each synapse develop independently from other neurons. so different synapses have different weights
What is the main difference between artificial and biological networks in terms of feature maps?
Biological: feature maps don’t represent the whole image with the same detail, strong bias towards sensitive/important parts. hands and eyes can move to sample relevant information
What is the main difference between artificial and biological networks in terms of filters?
Biological: filters have a limited spatial extent but not a fixed extent. multiple scales analysed at the same brain area ‘layer’. image layout is maintained in later stages, allowing analysis of spatial relationships at many levels
What is the main difference between artificial and biological networks in terms of filter weights?
Biological: filter weights are not shared across each layer, but similar filters are likely to emerge in many places.
What is the main difference between artificial and biological networks in terms of threshold?
Biological: maximum response is always part of the threshold
What is the main difference between artificial and biological networks in terms of normalisation?
Biological: activity is normalised by local mean activity rather than global mean activity
What is the main difference between artificial and biological networks in terms of learning?
Biological: learning is generally unsupervised
