Deep Dive 1 Flashcards

1
Q

Assumptions artificial intelligence and human mind

A
  • If we know how a computer generates knowledge and uses it, we will know how human memory works
  • If we know how human memory works, we can make a supercomputer
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

Collins & Quillian Model of semantic memory

A

It is a hierarchical network model. So it has categories which then have properties attached to them. Each category also has nodes underneath it which branch out. So e.g. bird which branches out to canary and ostrich.

RT depends on the distance in the network.

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

Criticisms of Collins and Quillian model

A
  • Theory has no explanation for typicality effects.
  • Frequency of association is moer important than distance in the hierarchy.
  • Also the theory can’t explain how NO-answers are created
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

Level of category organization

A
  • Superordinate (fruit)
  • Basic (appel), fastest processing and best remembered
  • Subordinate (honeycrisp)
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

Learning regularities

A

A good classification is reached after many examples (40 or more). This learning is quite difficult.
The prototype is typically misrecognized as seen, even if it wasn’t.
This suggests that subjects induce a sort of prototype.
Amnesia patients can also learn this type of categories.

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

Prototype theory

A

The closer to the prototype, the faster the response.
The prototype has no information on size of category or variability within category.

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

Exemplar theory

A

What is this? questions are answered by retrieving instances from memory, exemplars.
Retrieve all birds and compare them to be categorized item.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

Prototype vs exemplar theory

A

Exemplar theory considered best but makes the weird remember all assumption.
Neural network models are able to incorporate aspects of both.

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

Bartlett on semantic memory

A

He said memory is reconstructive. The contents of a memory are not preserved like an image on the computer. It is something that is reconstructed from bits and pieces of episodic memory which are fitted into semantic pattens (schemas).

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

Schema’s in semantic memory

A

Schema’s can distort recall of semantic memory. However they also help with encoding.

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

Scripts

A

A script is a schema with an explicit time ordering.
Originates in artificial intelligence.
Example: going to a restaurant. Has multiple steps that always happen.

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

Semantic dementia

A

Progressive loss of semantic knowledge.
Word-finding problems.
Comprehension difficulties.
No problem with new learning.
Lesions mainly located in infero-lateral temporal cortext
First loss of subordinate categories, superordinate survives longest

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