Download 5 levels of Neural Theory of Language

Survey
yes no Was this document useful for you?
   Thank you for your participation!

* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project

Document related concepts

Convolutional neural network wikipedia , lookup

Neural modeling fields wikipedia , lookup

Optogenetics wikipedia , lookup

Metastability in the brain wikipedia , lookup

Long-term depression wikipedia , lookup

State-dependent memory wikipedia , lookup

Holonomic brain theory wikipedia , lookup

Perceptual learning wikipedia , lookup

Neural engineering wikipedia , lookup

Nervous system network models wikipedia , lookup

Memory consolidation wikipedia , lookup

Node of Ranvier wikipedia , lookup

Synaptogenesis wikipedia , lookup

Neuropsychopharmacology wikipedia , lookup

Channelrhodopsin wikipedia , lookup

Nonsynaptic plasticity wikipedia , lookup

Eyeblink conditioning wikipedia , lookup

Development of the nervous system wikipedia , lookup

Artificial neural network wikipedia , lookup

Neuroanatomy of memory wikipedia , lookup

Learning wikipedia , lookup

Learning theory (education) wikipedia , lookup

Machine learning wikipedia , lookup

Epigenetics in learning and memory wikipedia , lookup

Catastrophic interference wikipedia , lookup

Concept learning wikipedia , lookup

Types of artificial neural networks wikipedia , lookup

Long-term potentiation wikipedia , lookup

Activity-dependent plasticity wikipedia , lookup

Hierarchical temporal memory wikipedia , lookup

Recurrent neural network wikipedia , lookup

Donald O. Hebb wikipedia , lookup

Transcript
20 Minute Quiz
For each of the questions, you can use text, diagrams,
bullet points, etc.
1)
2)
3)
4)
Give three processes that keep neural computation
from proceeding faster than at millisecond scale.
How does pre-natal activity dependent visual tuning
work?
How does the knee-jerk reflex work? Describe another
similar reflex found in humans.
Identify two philosophical issues discussed in the
assigned chapters of the M2M book.
How does activity lead to structural
change?


The brain (pre-natal, post-natal, and adult) exhibits a
surprising degree of activity dependent tuning and
plasticity.
To understand the nature and limits of the tuning and
plasticity mechanisms we study


How activity is converted to structural changes (say the ocular
dominance column formation)
It is centrally important


to arrive at biological accounts of perceptual, motor, cognitive
and language learning
Biological Learning is concerned with this topic.
Learning and Memory: Introduction
Learning and Memory
Declarative
Skill based
Fact based/Explicit
Implicit
Episodic
facts about a
situation
Semantic
general facts
Procedural
skills
Skill and Fact Learning may involve
different mechanisms

Certain brain injuries involving the hippocampal region of
the brain render their victims incapable of learning any
new facts or new situations or faces.


Fact learning can be single-instance based. Skill learning
requires repeated exposure to stimuli


But these people can still learn new skills, including
relatively abstract skills like solving puzzles.
subcortical structures like the cerebellum and basal ganglia
seem to play a role in skill learning
Implications for Language Learning?
Models of Learning
Hebbian ~ coincidence
 Recruitment ~ one trial
 Supervised ~ correction (backprop)
 Reinforcement ~ delayed reward
 Unsupervised ~ similarity

Hebb’s Rule


The key idea underlying theories of neural
learning go back to the Canadian
psychologist Donald Hebb and is called
Hebb’s rule.
From an information processing
perspective, the goal of the system is to
increase the strength of the neural
connections that are effective.
Hebb (1949)
“When an axon of cell A is near enough to
excite a cell B and repeatedly or
persistently takes part in firing it, some
growth process or metabolic change takes
place in one or both cells such that A’s
efficiency, as one of the cells firing B, is
increased”
From: The organization of behavior.
Hebb’s rule

Each time that a particular synaptic connection
is active, see if the receiving cell also becomes
active. If so, the connection contributed to the
success (firing) of the receiving cell and should
be strengthened. If the receiving cell was not
active in this time period, our synapse did not
contribute to the success the trend and should
be weakened.
LTP and Hebb’s Rule

Hebb’s Rule:
neurons that fire together wire together
strengthen
weaken
Long Term Potentiation (LTP) is the
biological basis of Hebb’s Rule
 Calcium channels are the key mechanism

Chemical realization of Hebb’s rule

It turns out that there are elegant chemical processes
that realize Hebbian learning at two distinct time scales



Early Long Term Potentiation (LTP)
Late LTP
These provide the temporal and structural bridge from
short term electrical activity, through intermediate
memory, to long term structural changes.
Calcium Channels Facilitate
Learning

In addition to the synaptic channels responsible
for neural signaling, there are also Calciumbased channels that facilitate learning.
 As
Hebb suggested, when a receiving neuron fires,
chemical changes take place at each synapse that
was active shortly before the event.
Long Term Potentiation (LTP)


These changes make each of the winning synapses
more potent for an intermediate period, lasting from
hours to days (LTP).
In addition, repetition of a pattern of successful firing
triggers additional chemical changes that lead, in time, to
an increase in the number of receptor channels
associated with successful synapses - the requisite
structural change for long term memory.

There are also related processes for weakening synapses and
also for strengthening pairs of synapses that are active at about
the same time.
LTP is found in the hippocampus
Essential for declarative memory
(Episodic Memory)
In the temporal lobe
Cylindrical Structure
Amygdala
Hippocampus
Temporal
lobe
The Hebb rule is found with long term
potentiation (LTP) in the hippocampus
Schafer collateral pathway
Pyramidal cells
1 sec. stimuli
At 100 hz
During normal
low-frequency
trans-mission,
glutamate
interacts with
NMDA and nonNMDA (AMPA)
and metabotropic
receptors.
With highfrequency
stimulation
Early and late LTP
(Kandel, ER, JH Schwartz and TM Jessell
(2000) Principles of Neural Science.
New York: McGraw-Hill.)
A.
Experimental setup for demonstrating
LTP in the hippocampus. The Schaffer
collateral pathway is stimulated to
cause a response in pyramidal cells of
CA1.
B.
Comparison of EPSP size in early and
late LTP with the early phase evoked
by a single train and the late phase by 4
trains of pulses.
Computational Models based on
Hebb’s rule
Many computational systems for modeling
incorporate versions of Hebb’s rule.
 Winner-Take-All:




Recruitment Learning


Units compete to learn, or update their weights.
The processing element with the largest output is declared the
winner
Lateral inhibition of its competitors.
Learning Triangle Nodes
LTP in Episodic Memory Formation
A possible computational
interpretation of Hebb’s rule
j
wij
i
How often when unit j was firing, was unit i also firing?
Wij = number of times both units i and j were firing
-----------------------------------------------------number of times unit j was firing
Extensions of the basic idea

Bienenstock, Cooper, Munro Model (BCM model).

Basic extensions: Threshold and Decay.

Synaptic weight change proportional to the post-synaptic activation
as long as the activation is above threshold





Less than threshold decreases w
Over threshold increases w
Absence of input stimulus causes the postsynaptic potential to
decrease (decay) over time.
Other models (Oja’s rule) improve on this in various
ways to make the rule more stable (weights in the range
0 to 1)
Many different types of networks including Hopfield
networks and Boltzman machines can be trained using
versions of Hebb’s rule
Winner take all networks (WTA)
Often use lateral inhibition
 Weights are trained using a variant of
Hebb’s rule.
 Useful in pruning connections

 such
as in axon guidance
WTA: Stimulus ‘at’ is presented
1
a
2
t
o
Competition starts at category level
1
a
2
t
o
Competition resolves
1
a
2
t
o
Hebbian learning takes place
1
a
2
t
o
Category node 2 now represents ‘at’
Presenting ‘to’ leads to activation of
category node 1
1
a
2
t
o
Presenting ‘to’ leads to activation of
category node 1
1
a
2
t
o
Presenting ‘to’ leads to activation of
category node 1
1
a
2
t
o
Presenting ‘to’ leads to activation of
category node 1
1
a
2
t
o
Category 1 is established through
Hebbian learning as well
1
a
2
t
o
Category node 1 now represents ‘to’
Connectionist Model of
Word Recognition (Rumelhart and
McClelland)
Recruiting connections

Given that LTP involves synaptic strength
changes and Hebb’s rule involves
coincident-activation based strengthening
of connections
 How
can connections between two nodes be
recruited using Hebbs’s rule?
X
Y
X
Y
Finding a Connection in Random Networks
For Networks with N nodes and N branching factor,
there is a high probability of finding good links.
(Valiant 1995)
Recruiting a Connection in Random Networks
Informal Algorithm
1. Activate the two nodes to be linked
2.
Have nodes with double activation
strengthen their active synapses (Hebb)
3. There is evidence for a “now print” signal
based on LTP (episodic memory)
Triangle nodes and recruitment
Posture
Push
Palm
Recruiting triangle nodes
WTA TRIANGLE NETWORK
CONCEPT UNITS
A
L
B
recruited
C
K
free
E
F
G
Has-color
Green
Has-shape
Round
Has-color
GREEN
Has-shape
ROUND
Hebb’s rule is insufficient
tastebud
tastes rotten
eats food
gets sick
drinks water

should you “punish” all the connections?
So what to do?

Reinforcement Learning
 Use
the reward given by the environment
 For every situation, based on experience
learn which action(s) to take such that on
average you maximize total expected reward
from that situation.

There is now a biological story for
reinforcement learning (later lectures).
Models of Learning
Hebbian ~ coincidence
 Recruitment ~ one trial
 Next Lecture: Supervised ~ correction
(backprop)
 Reinforcement ~ delayed reward
 Unsupervised ~ similarity

Constraints on Connectionist
Models
100 Step Rule
Human reaction times ~ 100 milliseconds
Neural signaling time ~ 1 millisecond
Simple messages between neurons
Long connections are rare
No new connections during learning
Developmentally plausible
5 levels of Neural Theory of
Language
Pyscholinguistic
experiments
Spatial
Relation
Motor
Control
Metaphor Grammar
Cognition and Language
abstraction
Computation
Structured Connectionism
Triangle Nodes
Neural Net and
learning
SHRUTI
Computational Neurobiology
Biology
Neural
Development
Quiz
Midterm
Finals