2. Issue #2: Shift Invariance
Backprop cannot handle shift invariance (it cannot generalize from 0011, 0110 to 1100)
But the cup is on the table whether you see it right in the center or from the corner of your eyes (i.e. in different areas of the retina map)
What structure can we utilize to make the input shift-invariant?

4. Topological Relations
Separation
Contact
Coincidence:
Overlap
Inclusion
Encircle/surround
For the encircle or surround relation, the two regions occupy different areas of space, but one region completely surrounds the other…
If you think of the blue region as a closed boundary, then it defines a bounded region of its own, and the yellow region is inside this bounded region but is not coincident with the blue region itself….

9. Limitations Scale Uniqueness/Plausibility Grammar Abstract Concepts
Inference
Representation

10. 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 for us to understand these mechanisms to arrive at biological accounts of perceptual, motor, cognitive and language learning
Biological Learning is concerned with this topic.

11. Learning and Memory: Introduction
Declarative
Non-Declarative
Episodic
Semantic
Procedural
Memory
memory of a situation
general facts
skills

12. Learning and Memory: Introduction
There are two different types of learning
Skill Learning
Fact and Situation Learning
General Fact Learning
Episodic Learning
There is good evidence that the process underlying skill (procedural) learning is partially different from those underlying fact/situation (declarative) learning.

13. Skill and Fact Learning 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.
But these people can still learn new skills, including relatively abstract skills like solving puzzles.
Fact learning can be single-instance based. Skill learning requires repeated exposure to stimuli.
Implications for Language Learning?

14. Short term memory
How do we remember someone’s telephone number just after they tell us or the words in this sentence?
Short term memory is known to have a different biological basis than long term memory of either facts or skills.
We now know that this kind of short term memory depends upon ongoing electrical activity in the brain.
You can keep something in mind by rehearsing it, but this will interfere with your thinking about anything else. (Phonological Loop)

15. Long term memory
But we do recall memories from decades past.
These long term memories are known to be based on structural changes in the synaptic connections between neurons.
Such permanent changes require the construction of new protein molecules and their establishment in the membranes of the synapses connecting neurons, and this can take several hours.
Thus there is a huge time gap between short term memory that lasts only for a few seconds and the building of long-term memory that takes hours to accomplish.
In addition to bridging the time gap, the brain needs mechanisms for converting the content of a memory from electrical to structural form.

16. Situational Memory
Think about an old situation that you still remember well. Your memory will include multiple modalities- vision, emotion, sound, smell, etc.
The standard theory is that memories in each particular modality activate much of the brain circuitry from the original experience.
There is general agreement that the Hippocampal area contains circuitry that can bind together the various aspects of an important experience into a coherent memory.
This process is believed to involve the Calcium based potentiation (LTP).

17. Models of Learning Hebbian ~ coincidence Recruitment ~ one trial
Supervised ~ correction (backprop)
Reinforcement ~ delayed reward
Unsupervised ~ similarity

18. 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.

19. LTP and Hebb’s Rule
Hebb’s Rule: neurons that fire together wire together
Long Term Potentiation (LTP) is the biological basis of Hebb’s Rule
Calcium channels are the key mechanism
strengthen
weaken

20. 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.

21. 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.

22. The Hebb rule is found with long term potentiation (LTP) in the hippocampus
Schafer collateral pathway
Pyramidal cells
1 sec. stimuli
At 100 hz

23. Computational Models based on Hebb’s rule
The activity-dependent tuning of the developing nervous system, as well as post-natal learning and development, do well by following Hebb’s rule.
Explicit Memory in mammals appears to involve LTP in the Hippocampus.
Many computational systems for modeling incorporate versions of Hebb’s rule.
Winner-Take-All:
Units compete to learn, or update their weights.
The processing element with the largest output is declared the winner
Lateral inhibition of its competitors.
Recruitment Learning
Learning Triangle Nodes
LTP in Episodic Memory Formation

24. Hebb’s rule is not sufficient
What happens if the neural circuit fires perfectly, but the result is very bad for the animal, like eating something sickening?
A pure invocation of Hebb’s rule would strengthen all participating connections, which can’t be good.
On the other hand, it isn’t right to weaken all the active connections involved; much of the activity was just recognizing the situation – we would like to change only those connections that led to the wrong decision.
No one knows how to specify a learning rule that will change exactly the offending connections when an error occurs.
Computer systems, and presumably nature as well, rely upon statistical learning rules that tend to make the right changes over time. More in later lectures.

25. Models of Learning Hebbian ~ coincidence Recruitment ~ one trial
Supervised ~ correction (backprop)
Reinforcement ~ delayed reward
Unsupervised ~ similarity

26. LTP and one-shot memory
Twin requirements of LTP induction:
presynaptic activity + postsynaptic depolarization:
LTP requires synchronous activity at multiple synapses of a postsynaptic cell (cooperativity)
ideal for transforming a transient synchronous- activity based expression of a relation between multiple items into a persistent synaptic-efficacy based encoding of the relation (Shastri, 2001)

27. 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?

28. The Idea of Recruitment LearningB N K
F = B/N
Suppose we want to link up node X to node Y
The idea is to pick the two nodes in the middle to link them up
Can we be sure that we can find a path to get from X to Y? X Y
the point is, with a fan-out of 1000,
if we allow 2 intermediate layers,
we can almost always find a path

29. # Paths = (1-P k-1)*(N/F) = (1-P k-1)*B
Finding a Connection
P = (1-F) **B**K
P = Probability of NO link between X and Y
N = Number of units in a “layer”
B = Number of randomly outgoing units per unit
F = B/N , the branching factor
K = Number of Intermediate layers, 2 in the example N= K=
.999
.9999
.99999 1
.367
.905
.989 2
10-440
10-44
10-5
# Paths = (1-P k-1)*(N/F) = (1-P k-1)*B

30. X Y

31. X Y

32. Finding a Connection in Random Networks
For Networks with N nodes and branching factor,
there is a high probability of finding good links.
(Valiant 1995)

33. Recruiting a Connection in Random Networks
Informal Algorithm
Activate the two nodes to be linked
Have nodes with double activation strengthen their active synapses (Hebb)
There is evidence for a “now print” signal based on LTP (episodic memory)

34. Triangle nodes and feature structuresB C A A B C

35. Representing concepts using triangle nodes

36. Recruiting triangle nodes
Let’s say we are trying to remember a green circle
currently weak connections between concepts (dotted lines)
has-color
has-shape
blue
green
round
oval

37. Strengthen these connections
and you end up with this picture
has-color
has-shape
Green circle
blue
green
round
oval

39. Has-color
Has-shape
Green
Round

40. Has-color
Has-shape
GREEN
ROUND

41. Models of Learning Hebbian ~ coincidence Recruitment ~ one trial
Supervised ~ correction (backprop)
Reinforcement ~ delayed reward, soon
Unsupervised ~ similarity

42. 5 levels of Neural Theory of Language
Pyscholinguistic experiments
Spatial Relation
Motor Control
Metaphor
Grammar
Cognition and Language
Computation
Structured Connectionism
abstraction
Neural Net and learning
SHRUTI
Triangle Nodes
Computational Neurobiology
Biology
Neural Development
Quiz
Midterm
Finals

43. Tinbergen’s Four Questions
How does it work?
How does it improve fitness?
How does it develop and adapt?
How did it evolve?

44. postsynaptic
neuron
science-education.nih.gov

45. Brains ~ Computers 1000 operations/sec 100,000,000,000 units
10,000 connections/
graded, stochastic
embodied
fault tolerant
evolves, learns
1,000,000,000 ops/sec
1-100 processors
~ 4 connections
binary, deterministic
abstract
crashes
designed, programmed

47. Artist’s rendition of a typical cell membrane

49. Flexor-
Crossed
Extensor
Reflex
(Sheridan
1900)
Reflex
Circuits
With
Inter-neurons
Painful Stimulus

50. Gaits of the cat: an informal computational model

51. Neural Tissue
The skin and neural tissue arise from a single layer, known as the ectoderm
in response to signals provided by an adjacent layer, known as the mesoderm.
A number of molecules interact to determine whether the ectoderm becomes neural tissue or develops in another way to become skin

53. Critical Periods in Development
There are critical periods in development (pre and post-natal) where stimulation is essential for fine tuning of brain connections.
Other examples of columns
Orientation columns

54. Pre-Natal Tuning: Internally generated tuning signals
But in the womb, what provides the feedback to establish which neural circuits are the right ones to strengthen?
Not a problem for motor circuits - the feedback and control networks for basic physical actions can be refined as the infant moves its limbs and indeed, this is what happens.
But there is no vision in the womb. Recent research shows that systematic moving patterns of activity are spontaneously generated pre-natally in the retina. A predictable pattern, changing over time, provides excellent training data for tuning the connections between visual maps.
The pre-natal development of the auditory system is also interesting and is directly relevant to our story.
Research indicates that infants, immediately after birth, preferentially recognize the sounds of their native language over others. The assumption is that similar activity-dependent tuning mechanisms work with speech signals perceived in the womb.

55. Post-natal environmental tuning
The pre-natal tuning of neural connections using simulated activity can work quite well –
a newborn colt or calf is essentially functional at birth.
This is necessary because the herd is always on the move.
Many animals, including people, do much of their development after birth and activity-dependent mechanisms can exploit experience in the real world.
In fact, such experience is absolutely necessary for normal development.
As we saw, early experiments with kittens showed that there are fairly short critical periods during which animals deprived of visual input could lose forever their ability to see motion, vertical lines, etc.
For a similar reason, if a human child has one weak eye, the doctor will sometimes place a patch over the stronger one, forcing the weaker eye to gain experience.

56. Learning Rule – Gradient Descent on an Root Mean Square (RMS)
Learn wi’s that minimize squared error
O = output layer

57. Backpropagation Algorithm
Initialize all weights to small random numbers
For each training example do
For each hidden unit h:
For each output unit k:
Update each network weight wij:
with

58. Distributed vs Localist Rep’n
John 1
Paul
George
Ringo
John 1
Paul
George
Ringo
What are the drawbacks of each representation?

59. Distributed vs Localist Rep’n
John 1
Paul
George
Ringo
John 1
Paul
George
Ringo
What happens if you want to represent a group?
How many persons can you represent with n bits? 2^n
What happens if one neuron dies?
How many persons can you represent with n bits? n

60. Word Superiority Effect

61. Modeling lexical access errors
Semantic error
Formal error (i.e. errors related by form)
Mixed error (semantic + formal)
Phonological access error

62. Phonological access error: Selection of incorrect phonemes
Syl
FOG
DOG
CAT
RAT
MAT
On Vo Co f r d k m ae o t g
Onsets
Vowels
Codas
Adapted from Gary Dell, “Producing words from pictures or from other words”

63. MRI and fMRI MRI: Images of brain structure.
fMRI: Images of brain function.
Tissues differ in magnetic susceptibility (grey matter, white matter, cerebrospinal fluid)

64. Mirror Neurons: Area F5
Mirror neurons were first discovered by Gallese and his collegues in Measuring neuronal firing in F5, a premotor area of the macaque brain, they found neuorns that were active both when the monkey made an action, and when the monkey saw someone else perform the same action. Thus, these neurons seem to code not only accross motor and visual domains, but also across self and other.
In 1996 Gallese and his collegues, while measuring neuronal activation in F5, a premotor area of the macaque brain, found neurons that not only were activated by the monkey producing a specific action, but when the monkey watched someone else perform the same action. These neurons were called mirror neurons. (refer to pic – so here you see…) A subgroup of these neurons were also active across multiple motor effectors of the action – hence they were activated by actions with the left hand, with the right hand, the same action with the mouth. Mirror neurons there seem to code for goal oriented actions, regardless of both the doer, and to some degree, the means by which the action is performed

65. Cortical Mechanism for Action Recognition
adds additional somatosensory information to
the movement to be imitated
provides an early
description of the action
Observed
Action
STS
Parietal mirror neurons (PF)
(inferior parietal lobule)
Frontal mirror neurons (F5) (BA 44)
copies of the motor plans necessary to imitate actions for monitoring purposes
F-p share lots of info, both can provide efference copies to sts
codes the goal of the
action to be imitated

66. Somatotopy of Action Observation
Foot Action
Hand Action
Mirror Neurons: Parietal, Premotor (Area ), and Broca’s (45)
A: mimicked actions; no parietal activation
B: Action on objects; Parietal activation.
Point: location determined by action.
Mouth Action
Buccino et al. Eur J Neurosci 2001

67. The WCS Color Chips Basic color terms: Single word (not blue-green)
Frequently used (not mauve)
Refers primarily to colors (not lime)
Applies to any object (not blonde)
FYI:
English has 11 basic color terms

68. Concepts are not categorical