# US 2014/0156578 A1 · Firing rate independent spike message passing in large scale neural network modeling

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2014/0156578 A1
- Document ID: US20140156578A1
- Status: Published application
- Kind: A1
- Application number: 14/094,589
- Filed: 2013-12-02
- Published: 2014-06-05
- Earliest claimed: 2012-12-03
- Assignees: HRL Laboratories, LLC
- Applicants: Hrl Laboratories, LLC
- Inventors: Corey Thibeault; Kirill Minkovich; Narayan Srinivasa
- Web page: https://patentgrep.com/patent/US20140156578A1

## CPC

- G06N 3/049: Temporal neural networks, e.g. delay elements, oscillating neurons or pulsed inputs
- G06N 3/063: using electronic means
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS

## Abstract

A neural network portion comprising N pre-synaptic neurons capable each of firing an action potential, wherein the number N can be encoded in a word of n bits; the neural network portion being provided for, upon firing of a number F of pre-synaptic neurons in a predetermined period of time: if F.n<N, generating a first type message, the message comprising a unique address for each pre-synaptic neuron having fired in said predetermined period of time, each address being encoded as a word of n bits; and if F.n>N, generating a second type message, the message comprising N bits and being encoded in words of n bits, wherein each one of said N pre-synaptic neurons is represented by a unique bit, each bit having a first value if the pre-synaptic neuron represented by the bit fired in said predetermined period of time, and a second value otherwise.

## Claims

### Claim 1

A neural network portion comprising a number N of pre-synaptic neurons capable each of firing an action potential, wherein N can be encoded in a word of n bits;
the neural network portion being provided for, upon firing of a number F of pre-synaptic neurons in a predetermined period of time:
if F.n<N, generating a message of a first type, the message comprising a unique address for each pre-synaptic neuron having fired in said predetermined period of time, each address being encoded as a word of n bits; and
if F.n>N, generating a message of a second type, the message comprising N bits and being encoded in words of n bits, wherein each one of said N pre-synaptic neurons is represented by a unique bit, each bit having a first value if the pre-synaptic neuron represented by the bit has fired in said predetermined period of time, and a second value otherwise.

### Claim 2 (depends on 1)

The neural network portion of claim 1, wherein 2n-1<N<=2n.

### Claim 3 (depends on 1)

The neural network portion of claim 1, further comprising a post synaptic neuron associated to at least some of the N pre-synaptic neurons, the neural network portion being provided to decode either of said message of the first type and said message of the second type and, in response to said decoding, send to the post synaptic neuron information of which one of the N pre-synaptic neurons associated to the post-synaptic neuron have fired in said predetermined period of time;
the pre-synaptic neurons being arranged in a first node of the neural network portion and the post-synaptic neuron being arranged in a second node of the neural network portion.

### Claim 4 (depends on 1)

The neural network portion of claim 1, further comprising a number N of post synaptic neurons associated each to a single pre-synaptic neuron, the neural network portion being provided to decode either of said message of the first type and said message of the second type, and in response to said decoding to cause the firing of each post synaptic neuron associated to a pre-synaptic neuron that fired in said predetermined period of time;
the pre-synaptic neurons being arranged in a first node of the neural network portion and the post-synaptic neurons being arranged in a second node of the neural network portion.

### Claim 5 (depends on 3)

A neural network system comprising:
a number p of neural network portions according to claim 3, comprising each a group of Ni pre-synaptic neurons, with i=1 to p.

### Claim 6 (depends on 5)

The neural network system of claim 5, wherein the post-synaptic neurons associated with each group of pre-synaptic neurons are distinct.

### Claim 7 (depends on 6)

The neural network system of claim 6, wherein at least two of the groups of Ni pre-synaptic neurons share at least one common pre-synaptic neuron.

### Claim 8 (depends on 3)

The neural network portion of claim 3, wherein the neural network portion is provided to synchronize the messages of the first and second types as if the post and pre-synaptic neurons were located on a same computer node.

### Claim 9 (depends on 1)

The neural network portion of claim 1, wherein said messages of the first and second type comprise each a header indicating the type of the message.

### Claim 10 (depends on 9)

The neural network portion of claim 9, wherein the header of said message of the first type further indicates the number F.

### Claim 11 (depends on 3)

The neural network portion of claim 3, wherein the sensitivity of the post synaptic neuron to the firing of at least some of the pre-synaptic neurons is plastic.

### Claim 12

A method of transmitting firing information in a neural network portion comprising a number N of pre-synaptic neurons capable each of firing an action potential, wherein N can be encoded in a word of n bits;
the method comprising, upon firing of a number F of pre-synaptic neurons in a predetermined period of time:
if F.n<N, generating a message of a first type, the message comprising a unique address for each pre-synaptic neuron having fired in said predetermined period of time, each address being encoded as a word of n bits; and
if F.n>N, generating a message of a second type, the message comprising N bits and being encoded in words of n bits, wherein each one of said N pre-synaptic neurons is represented by a unique bit, each bit having a first value if the pre-synaptic neuron represented by the bit has fired in said predetermined period of time, and a second value otherwise.

### Claim 13 (depends on 12)

The method of claim 12, wherein 2n-1<N<=2n.

### Claim 14 (depends on 12)

The method of claim 12, wherein the neural network portion comprises a post synaptic neuron associated to at least some of the N pre-synaptic neurons; the pre-synaptic neurons being arranged in a first node of the neural network portion and the post-synaptic neuron being arranged in a second node of the neural network portion; the method comprising:
decoding either of said message of the first type and said message of the second type; and
in response to said decoding, sending to the post synaptic neuron information about which ones of the N pre-synaptic neurons associated to the post-synaptic neuron have fired in said predetermined period of time.

### Claim 15 (depends on 12)

The method of claim 12, wherein the neural network portion comprises a number N of post synaptic neurons associated each to a single pre-synaptic neuron; the pre-synaptic neurons being arranged in a first node of the neural network portion and the post-synaptic neurons being arranged in a second node of the neural network portion the method comprising:
decoding either of said message of the first type and said message of the second type; and
in response to said decoding, cause the firing of each post synaptic neuron associated to a pre-synaptic neuron that fired in said predetermined period of time.

### Claim 16 (depends on 14)

A method of transmitting firing information in a neural network system, the neural network system comprising a number p of neural network portions comprising each a group of Ni pre-synaptic neurons, with i=1 to p; the pre-synaptic neurons of each group being associated to a post-synaptic neuron; the method comprising:
generating a message of the first or second type according to the method of claim 14 for each neural network portion of the neural network system.

### Claim 17 (depends on 16)

The method of claim 16, wherein each group of pre-synaptic neurons is associated to a distinct post-synaptic neuron.

### Claim 18 (depends on 17)

The method of claim 17, wherein at least two of the groups of Ni pre-synaptic neurons share at least one common pre-synaptic neuron.

### Claim 19 (depends on 14)

The method of claim 14, further comprising synchronizing the messages of the first and second types as if the post and pre-synaptic neurons were located on a same computer node.

### Claim 20 (depends on 12)

The method of claim 12, comprising providing in said messages of the first and second type a header indicating the type of the message.

### Claim 21 (depends on 20)

The method of claim 20, comprising indicating the number F in the header of said message of the first type.

### Claim 22 (depends on 14)

The method of claim 14, wherein the sensitivity of the post synaptic neuron to the firing of at least some of the pre-synaptic neurons is plastic.