Reading...
Play button
Play button
Use LEFT and RIGHT arrow keys to navigate between flashcards;
Use UP and DOWN arrow keys to flip the card;
H to show hint;
A reads text to speech;
60 Cards in this Set
- Front
- Back
|
activation unit
|
become active in certain environmental events
Used in Connectionist Models |
|
|
afterimage
|
the way the eyes process information and reacts to images
|
|
|
backward masking
|
A stimulus that came on afterward masks what came on before.
|
|
|
bottom-up processes
|
aka data driven processes.
Processing small bits of data to form meaning. We do not simply perceive data sets. |
|
|
cognitive demon
|
Detect objects pattern based on features
|
|
|
computational demon
|
Detect features based on element matching
|
|
|
conceptually-driven processes
|
High level reasoning
|
|
|
connectionist model
|
The central principle is that mental phenomena can be described by interconnected networks of simple units.
|
|
|
data demons
|
Detect the make-up of a feature.
Encodes sensory patters (feature elements) |
|
|
data-driven processes
|
aka bottom-up processes
|
|
|
decay
|
to decrease usually gradually in size, quantity, activity, or force
Sperling's Iconic memory matrix. |
|
|
decision demon
|
Retrieve meaning of pattern
|
|
|
dorsal “where” stream
|
Identifies where an object is in space
|
|
|
echoic memory
|
Auditory sensory memory
|
|
|
erasure
|
the visual cue erases (or merges) with the letter
(Averbach and Coriell matrix with underline and grid circle) |
|
|
feature theory
|
Detecting features of patterns.
We recognize simple features of a pattern. |
|
|
foveal vision
|
Does not exhibit lateral inhibition.
The reason why there is 100% reporting for the letters in the middle of the matrix |
|
|
geon
|
3-D primatives
Used to construct whole physical objects |
|
|
hidden layer
|
in between the input and output layer
A processing layer. |
|
|
iconic memory
|
Visual persistence
a pattern of visual information persists for a short duration |
|
|
input layer
|
A bunch of units that respond to an external stimulus
|
|
|
recognition by components
|
Biderman
Recognition occurs with successful decomposition to primitives |
|
|
lateral inhibition
|
a sensory limitation, not memory inhibition.
The reason why participants did not score 100% on the 2x12 matrix |
|
|
link weight
|
Some links count more than others.
Heavier weight - connect to more units. Determine amount of activation that is transmitted |
|
|
modality effect
|
superior end-of-list recall when items presented in auditory mode.
|
|
|
negative after image
|
The negatives of the true colors.
red - green blue - yellow |
|
|
non-recoverable objects
|
Objects that had a large amount of the drawing missing.
Had 25% correct guess rate Biderman researched the locus of deletion in an attempt to see whether vertices or midsegments led to better retrieval. |
|
|
non-selective readout
|
Randomly passing information into memory
ie - Sperling flashing matrix Occurs before selective readout |
|
|
output layer
|
Corresponds to the response of a system
|
|
|
pandemonium
|
Demons are the processing units of the mind.
Proposed by Selfridge A form of feature approach |
|
|
partial report
|
A tone tells which row of letters the participant needs to report.
Performance - all three items are consistently reported. Supports the notion of large capacity of iconic memory |
|
|
perception
|
attaching meaning to patterns of sensations
|
|
|
precategorical
|
A term used to identify stimuli that have not had meaning attached to it.
|
|
|
preconscious
|
Do not have awareness of sensory inputs
|
|
|
receptive field
|
A stimulus to which a neuron response is maximal
Mostly involved with visual processing. Provides support that we are sensitive to features in our cognitive world. |
|
|
report cues
|
The cues used by researchers to indicate which row they need to repeat.
The longer the cue takes to stimulate, the less the person is able to repeat what he has seen. |
|
|
selective readout
|
sensory information is guided into memory by consciousness
|
|
|
recoverable objects
|
Objects that have recognizable features
Have a small percentage of degregation. |
|
|
RBC
|
Basic components that are used to construct whole physical objects are geons.
Geons = 3-D primatives |
|
|
sensation
|
the detection of physical energy that impinges on an organism
Not constantly aware of this |
|
|
sensory memory
|
Not conscious memory
|
|
|
sensory register
|
Characteristics:
Large Capacity Brief duration Veridical format Preconscious One per sense modality |
|
|
suffix effect
|
interference by end-of-list signal greater when word used compared to when tone used.
Why? Spoken word pushes the other word after it out of memory. |
|
|
Precategorical
|
Veridical format
Has no meaning attached to the stimuli |
|
|
template theory
|
For any particular pattern of sensory data, you match it with prior data and look for the meaning
Cumbersome theory. |
|
|
unit link
|
Fully interconnected with input layer, hidden layer and outer layer.
- They are weighted |
|
|
ventral “what” stream
|
Identifies what an object is
|
|
|
veridical format
|
Precategorical
A raw pattern of sensory input Has no meaning attached to it. |
|
|
visual agnosia
|
Two forms: Apperceptive and Associative
The inability to perceive objects. |
|
|
whole report
|
Used by Sperling.
Asked the participant to identify all of the letters that appeared in the matrix |
|
|
Apperceptive agnosia
|
Not able to perceive object at all
|
|
|
Associative
|
Not able to identify object
|
|
|
three-eared man technique
|
Presented three different pieces of sound info so that it sounds like it came from different locations.
Both sides - experienced as central location Proved that echoic memory is twelve times as long as iconic memory. |
|
|
backpropagation
|
Adjustment of the weights through feedback.
|
|
|
Delta rule
|
Rule that is applied to determine how much we change the weights
|
|
|
Training set
|
Show similar patterns and adjust its accuracy
looking for ~95% |
|
|
neural network
|
aka connectionist models.
How neurons interact and communicate with each other |
|
|
PDP models
parallel distribution processing |
Flow of activation through the links in parallel, in a distributed way
|
|
|
Beauty of connectionist Model
|
it learns what it knows, it is not programmed in.
|
|
|
Graceful degregation
|
In connectionist model - can remove information and continues to operate without impairment.
Akin to our brain. |
