An analog Swish gadget is implemented using one or more thermodynamic chips (neuro-thermodynamic processors). The analog Swish gadget takes a thermodynamic input and calculates a result of the Swish function thermodynamically according to an engineered potential used for oscillators and oscillator couplings for a set of oscillators that implement the analog Swish gadget. The analog Swish gadget returns the result of the Swish function as a thermodynamic output that may be relayed to other energy-based models of a thermodynamic computer. The input, processing, and output are all performed thermodynamically (e.g., in an analog fashion) without a need to convert the information into a classical representation (e.g., a classical representation such as information stored in memory of a classical computing device).
An analog Swish gadget is implemented using one or more thermodynamic chips (neuro-thermodynamic processors). The analog Swish gadget takes a thermodynamic input and calculates a result of the Swish function thermodynamically according to an engineered potential used for oscillators and oscillator couplings for a set of oscillators that implement the analog Swish gadget. The analog Swish gadget returns the result of the Swish function as a thermodynamic output that may be relayed to other energy-based models of a thermodynamic computer. The input, processing, and output are all performed thermodynamically (e.g., in an analog fashion) without a need to convert the information into a classical representation (e.g., a classical representation such as information stored in memory of a classical computing device).
A method of configuring circuits for generating random walks on a graph comprising vertices interconnected by edges comprises: determining a number of colors associated with the graph, wherein each edge connected to a respective vertex is associated with a different respective color; arranging probabilistic circuit modules (PCMs), wherein each PCM comprises first and second inputs, first and second outputs, and is associated with an edge; arranging pluralities of input and output nodes; connecting each output of each PCM associated with a first color to an output node; connecting each input of each PCM associated with a second color to an input node; and connecting each output to a PCM input or to an output node such that the PCM outputs associated with a respective color are each connected to different respective PCM inputs associated with a different color or to an output node.
4.
CONFIGURING CIRCUITS FOR GENERATING SAMPLES BASED ON PREDETERMINED FUNCTIONS
A method comprises: arranging a configurable circuit module comprising a plurality of input nodes, a plurality of output nodes, and a plurality of probabilistic circuit modules defining a different respective mapping from each input node to a different respective output node, wherein each probabilistic circuit module of the plurality of probabilistic circuit modules comprises a first input configured to receive a bias voltage; receiving a function; determining a respective bias voltage for each probabilistic circuit module of the plurality of probabilistic circuit modules based at least in part on the function; providing a respective voltage to each input node of the plurality of input nodes; and generating, by the configurable circuit module, a sample from a probability distribution at the plurality of output nodes, based at least on the respective bias voltages and the respective voltages provided to each input node.
H03K 19/177 - Logic circuits, i.e. having at least two inputs acting on one outputInverting circuits using specified components using elementary logic circuits as components arranged in matrix form
A thermodynamic selection of experts energy-based model gadget includes a SoftMax gadget, a set of oscillators having a potential which is used to modify input data, and multiple energy-based models for processing data. The SoftMax gadget may produce one-hot encoded vectors which may be used by the set of oscillators having a potential which is used to modify input data, such that the modified input data corresponds to one of the multiple energy-based models for processing data.
A method of configuring circuits for generating random walks on a graph comprising vertices interconnected by edges comprises: determining a number of colors associated with the graph, wherein each edge connected to a respective vertex is associated with a different respective color; arranging probabilistic circuit modules (PCMs), wherein each PCM comprises first and second inputs, first and second outputs, and is associated with an edge; arranging pluralities of input and output nodes; connecting each output of each PCM associated with a first color to an output node; connecting each input of each PCM associated with a second color to an input node; and connecting each output to a PCM input or to an output node such that the PCM outputs associated with a respective color are each connected to different respective PCM inputs associated with a different color or to an output node.
H03K 17/687 - Electronic switching or gating, i.e. not by contact-making and -breaking characterised by the use of specified components by the use, as active elements, of semiconductor devices the devices being field-effect transistors
G06F 30/327 - Logic synthesisBehaviour synthesis, e.g. mapping logic, HDL to netlist, high-level language to RTL or netlist
8.
CONFIGURING CIRCUITS FOR GENERATING SAMPLES BASED ON PREDETERMINED FUNCTIONS
A method comprises: arranging a configurable circuit module comprising a plurality of input nodes, a plurality of output nodes, and a plurality of probabilistic circuit modules defining a different respective mapping from each input node to a different respective output node, wherein each probabilistic circuit module of the plurality of probabilistic circuit modules comprises a first input configured to receive a bias voltage; receiving a function; determining a respective bias voltage for each probabilistic circuit module of the plurality of probabilistic circuit modules based at least in part on the function; providing a respective voltage to each input node of the plurality of input nodes; and generating, by the configurable circuit module, a sample from a probability distribution at the plurality of output nodes, based at least on the respective bias voltages and the respective voltages provided to each input node.
H03K 17/687 - Electronic switching or gating, i.e. not by contact-making and -breaking characterised by the use of specified components by the use, as active elements, of semiconductor devices the devices being field-effect transistors
9.
THERMODYNAMIC COMPUTING SYSTEM CONFIGURED TO USE NATURAL GRADIENT DESCENT TECHNIQUES TO DETERMINE UPDATED WEIGHTS AND BIASES
A neuro-thermodynamic computer includes a thermodynamic chip that includes oscillators that are mapped to neurons and additional oscillators that are mapped to synapses, wherein the synapses correspond to weights and bias values used to describe relationships between the neurons in an energy-based model. Learning algorithms are described for computing gradients for a positive phase term and a negative phase term, as well as elements of an information matrix based on measurements taken of the synapse oscillators, without a need to fully compute updated weights and biases on classical hardware. However, classical hardware may be used to perform basic operations to convert the measured values into calculated updated weights and biases. The updated weights and bias values are used to train the energy-based model, which once trained, can be used to generate inferences for various types of machine learning or AI-type problems.
A method comprises: producing, using a first metastable circuit, a bistable state that varies over time between a first stable voltage and a second stable voltage, where a fraction of time that the bistable state spends at the first stable voltage is associated with a first probability; producing, using a first noise circuit, a first voltage distribution; producing, using a second noise circuit, a second voltage distribution; and producing, using a first mixer circuit, a third voltage distribution that is based at least in part on the bistable state, the first voltage distribution, and the second voltage distribution.
11.
CONFIGURING A CIRCUIT FOR GENERATING SAMPLES FROM A TARGET DISTRIBUTION
A method for configuring a circuit for generating samples from a target distribution comprises: receiving a matrix representing parameters associated with the target distribution; tuning a plurality of tunable capacitance circuits in a tunable capacitance network based at least in part on respective elements of the matrix, wherein the tunable capacitance network consists essentially of interconnected wires intersecting at a plurality of nodes with selected pairs of nodes of the plurality of nodes interconnected by a respective tunable capacitance circuit of the plurality of tunable capacitance circuits and one or more nodes of the plurality of nodes connected to a common ground by a respective tunable capacitance circuit of the plurality of tunable capacitance circuits; recording respective voltage samples from the plurality of nodes of the tunable capacitance network; and storing a linear transformation of a vector of the voltage samples based at least in part on the matrix.
G01R 19/28 - Arrangements for measuring currents or voltages or for indicating presence or sign thereof adapted for measuring in circuits having distributed constants
12.
CIRCUITS FOR PROVIDING COMBINATIONS OF INPUTS ASSOCIATED WITH BISTABLE STATES
An apparatus comprises: a metastable circuit module including more than two output nodes, wherein the more than two output nodes are associated with a bistable state that varies over time between a first and second stable voltage, where a fraction of time that the bistable state spends at the first stable voltage is associated with a probability; a plurality of noise circuits, where each noise circuit is configured to produce a voltage distribution; and a mixer circuit comprising more than two gate circuits, where each gate circuit is connected to a respective output node of the metastable circuit module and to a respective noise circuit; wherein the mixer circuit is configured to produce a voltage distribution that is based at least in part on each probability associated with the metastable circuit module and each voltage distribution associated with a respective noise circuit of the plurality of noise circuits.
H10D 80/30 - Assemblies of multiple devices comprising at least one device covered by this subclass the at least one device being covered by groups , e.g. assemblies comprising integrated circuit processor chips
13.
MANAGING ARRANGEMENTS OF TRANSISTORS IN CIRCUITS FOR GENERATING SAMPLES FROM A TARGET DISTRIBUTION
An apparatus comprises: one or more voltage sources; a plurality of transistor rails, each transistor rail of the plurality of transistor rails comprising a first transistor connected to a first node and a second node, a second transistor connected to the first node, and a first capacitor connected to the first node and the second node; wherein the first node of each transistor rail of the plurality of transistor rails is connected to a first node of two different respective transistor rails of the plurality of transistor rails by two different respective capacitors; wherein one or more voltage sources are configured to apply a respective set of voltages to each transistor rail of the plurality of transistor rails.
H10D 80/20 - Assemblies of multiple devices comprising at least one device covered by this subclass the at least one device being covered by groups , e.g. assemblies comprising capacitors, power FETs or Schottky diodes
14.
LEARNING WITH PARAMETERIZED QUANTUM-THERMODYNAMIC CIRCUITS
Systems and methods of preparing a mixed quantum state using a thermodynamic - quantum hybrid scheme are provided. A system comprises a thermodynamic processor, a quantum processor, and a controller operatively coupled to the thermodynamic processor and the quantum processor. A plurality of samples from a data state is received. A plurality of configuration parameters of a thermodynamic processor is initialized to initial values, the thermodynamic processor having a plurality of visible nodes and a plurality of hidden nodes. A plurality of configuration parameters of a quantum processor is initialized to initial values. The hidden nodes of the thermodynamic processor are sampled. Based on samples of the visible nodes and sample of the hidden nodes and on the data state, the configuration parameters of the thermodynamic processor are updated. Based on samples of the visible nodes and samples of the hidden nodes, the configuration parameters of the quantum processor are updated.
H03K 19/195 - Logic circuits, i.e. having at least two inputs acting on one outputInverting circuits using specified components using superconductive devices
G06N 3/06 - Physical realisation, i.e. hardware implementation of neural networks, neurons or parts of neurons
An apparatus comprises: one or more voltage sources; a plurality of transistor rails, each transistor rail of the plurality of transistor rails comprising a first transistor connected to a first node and a second node, a second transistor connected to the first node, and a first capacitor connected to the first node and the second node; wherein the first node of each transistor rail of the plurality of transistor rails is connected to a first node of two different respective transistor rails of the plurality of transistor rails by two different respective capacitors; wherein one or more voltage sources are configured to apply a respective set of voltages to each transistor rail of the plurality of transistor rails.
G05F 1/59 - Regulating voltage or current wherein the variable actually regulated by the final control device is DC using semiconductor devices in series with the load as final control devices including plural semiconductor devices as final control devices for a single load
G01R 31/26 - Testing of individual semiconductor devices
G05F 1/565 - Regulating voltage or current wherein the variable actually regulated by the final control device is DC using semiconductor devices in series with the load as final control devices sensing a condition of the system or its load in addition to means responsive to deviations in the output of the system, e.g. current, voltage, power factor
16.
CIRCUITS FOR PROVIDING COMBINATIONS OF INPUTS ASSOCIATED WITH BISTABLE STATES
An apparatus comprises: a metastable circuit module including more than two output nodes, wherein the more than two output nodes are associated with a bistable state that varies over time between a first and second stable voltage, where a fraction of time that the bistable state spends at the first stable voltage is associated with a probability; a plurality of noise circuits, where each noise circuit is configured to produce a voltage distribution; and a mixer circuit comprising more than two gate circuits, where each gate circuit is connected to a respective output node of the metastable circuit module and to a respective noise circuit; wherein the mixer circuit is configured to produce a voltage distribution that is based at least in part on each probability associated with the metastable circuit module and each voltage distribution associated with a respective noise circuit of the plurality of noise circuits.
H03K 19/20 - Logic circuits, i.e. having at least two inputs acting on one outputInverting circuits characterised by logic function, e.g. AND, OR, NOR, NOT circuits
17.
CONFIGURING A CIRCUIT FOR GENERATING SAMPLES FROM A TARGET DISTRIBUTION
A method for configuring a circuit for generating samples from a target distribution comprises: receiving a matrix representing parameters associated with the target distribution; tuning a plurality of tunable capacitance circuits in a tunable capacitance network based at least in part on respective elements of the matrix, wherein the tunable capacitance network consists essentially of interconnected wires intersecting at a plurality of nodes with selected pairs of nodes of the plurality of nodes interconnected by a respective tunable capacitance circuit of the plurality of tunable capacitance circuits and one or more nodes of the plurality of nodes connected to a common ground by a respective tunable capacitance circuit of the plurality of tunable capacitance circuits; recording respective voltage samples from the plurality of nodes of the tunable capacitance network; and storing a linear transformation of a vector of the voltage samples based at least in part on the matrix.
A method comprises: producing, using a first metastable circuit, a bistable state that varies over time between a first stable voltage and a second stable voltage, where a fraction of time that the bistable state spends at the first stable voltage is associated with a first probability; producing, using a first noise circuit, a first voltage distribution; producing, using a second noise circuit, a second voltage distribution; and producing, using a first mixer circuit, a third voltage distribution that is based at least in part on the bistable state, the first voltage distribution, and the second voltage distribution.
Systems, methods and computer readable media relating to a neuro-thermodynamic computers configured to train a learning model based on values representing gradient terms stored in position degrees of freedom of relay oscillators are described. An energy-based model comprising oscillators representing neurons and oscillators representing synapse values may be trained using gradient terms obtained in an analogue way. The gradient terms may be stored on respective relay oscillators and gradient terms may be combined with other gradient terms. Oscillators representing synapse parameters may be updated based on one or more gradient terms. In some embodiments, the training protocol is implemented in a fully analogue way. In some embodiments, measurements of relay oscillators are performed and stored in a classical computing device for post-processing.
A thermodynamic computing chip that is configured to emulate deep neural diffusion of a deep energy-based model (EBM) and update parameters of an energy function using diffusion recovery likelihood is disclosed. In some embodiments, a deep EBM may comprise one or more EBMs that process thermodynamic information via thermodynamic evolution. Relay oscillators or measurements may be utilized to obtain gradients of the deep EBM and sampled input values used to update parameters of the energy function.
A thermodynamic computing chip that is configured emulate deep neural diffusion of a deep energy-based model (EBM) and sample input values. In some embodiments, a deep EBM may comprise one or more EBMs that process thermodynamic information via thermodynamic evolution. Relay oscillators or measurements may be utilized to obtain gradients of the deep EBM and thus sample input values.
A thermodynamic computing chip that is configured to emulate deep neural diffusion of a deep energy-based model (EBM) and update parameters of an energy function using diffusion recovery likelihood is disclosed. In some embodiments, a deep EBM may comprise one or more EBMs that process thermodynamic information via thermodynamic evolution. Relay oscillators or measurements may be utilized to obtain gradients of the deep EBM and sampled input values used to update parameters of the energy function.
Systems, methods and computer readable media relating to a neuro-thermodynamic computers configured to train a learning model based on values representing gradient terms stored in position degrees of freedom of relay oscillators are described. An energy-based model comprising oscillators representing neurons and oscillators representing synapse values may be trained using gradient terms obtained in an analogue way. The gradient terms may be stored on respective relay oscillators and gradient terms may be combined with other gradient terms. Oscillators representing synapse parameters may be updated based on one or more gradient terms. In some embodiments, the training protocol is implemented in a fully analogue way. In some embodiments, measurements of relay oscillators are performed and stored in a classical computing device for post-processing.
A thermodynamic computing chip that is configured emulate deep neural diffusion of a deep energy-based model (EBM) and sample input values. In some embodiments, a deep EBM may comprise one or more EBMs that process thermodynamic information via thermodynamic evolution. Relay oscillators or measurements may be utilized to obtain gradients of the deep EBM and thus sample input values.
G06F 30/27 - Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
A thermodynamic relay gadget includes a set of one or more relay oscillators, an additional relay oscillator, a bias oscillator, and an on-chip controller. Respective ones of the relay oscillators have a time dependent mass or a time dependent frequency that is controllable, by the on-chip controller. The relay gadget is configured to relay thermodynamic information in analog form between an input oscillator of a first energy-based model and an output oscillator of a second energy-based model.
A thermodynamic relay gadget includes a set of one or more relay oscillators, an additional relay oscillator, a bias oscillator, and an on-chip controller. Respective ones of the relay oscillators have a time dependent mass or a time dependent frequency that is controllable, by the on-chip controller. The relay gadget is configured to relay thermodynamic information in analog form between an input oscillator of a first energy-based model and an output oscillator of a second energy-based model.
H03B 15/00 - Generation of oscillations using galvano-magnetic devices, e.g. Hall-effect devices, devices using spin transfer effects, devices using giant magnetoresistance, or using super-conductivity effects
A thermodynamic relay gadget includes a relay oscillator and an on-chip controller. The relay oscillator has a time dependent mass or time dependent frequency that is controllable, by the on-chip controller. The relay gadget is configured to relay thermodynamic information in analog form between an output oscillator of a first energy-based model and an input oscillator of a second energy-based model.
H03K 19/195 - Logic circuits, i.e. having at least two inputs acting on one outputInverting circuits using specified components using superconductive devices
28.
THERMODYNAMIC COMPUTING SYSTEM CONFIGURED TO IMPLEMENT LAYER NORMALIZATION ARCHITECTURE
Systems, methods and computer readable media relating to neuro-thermodynamic computers configured to implement a layer normalization gadget, wherein the layer normalization gadget is configured to perform layer normalization operations. Thermodynamic data may be used as input to one or more thermodynamic chips comprising oscillators, wherein thermodynamic evolution according to one or more energy potentials governing the oscillators enable results of layer normalization to be obtained by respective ones of the oscillators. Furthermore, the results may be encoded as thermodynamic data in position degree of freedoms of respective oscillators.
Systems, methods and computer readable media relating to neuro-thermodynamic computers configured to implement hierarchical architecture, wherein the hierarchical architecture includes one or more layers of components, and wherein the hierarchical architecture is configured to perform Gibbs sampling and nested Gibbs sampling. For example, a block layer may include an energy based model (EBM) implemented using oscillators and couplings between oscillators. A chip layer may include multiple blocks coupled to each other using relay oscillators. A package layer may include multiple chips coupled to each other using additional relay oscillators.
A thermodynamic mixture of experts gadget includes a SoftMax gadget, multiple energy-based models for processing data, and a summation gadget, also called a Selection of Experts gadget. The SoftMax gadget generates one-hot encoded vectors, which correspond to particular ones of the energy-based models for processing data. The outputs of the energy-based models for processing data, in combination with the one-hot encoded vectors, are inputs to the summation gadget, which generates output that is processed data, processed by energy-based models selected by the SoftMax gadget.
An analog SoftMax gadget is implemented using one or more thermodynamic chips (neuro-thermodynamic processors). The analog SoftMax gadget takes a thermodynamic input and calculates a result of the SoftMax function thermodynamically according to an engineered potential used for oscillators and oscillator couplings for a set of oscillators that implement the analog SoftMax gadget. The analog SoftMax gadget returns the result of the SoftMax function as a thermodynamic output that may be relayed to other energy-based models of a thermodynamic computer. The input, processing, and output are all performed thermodynamically (e.g., in an analog fashion) without a need to convert the information into a classical representation.
Systems, methods and computer readable media relating to neuro-thermodynamic computers configured to implement a layer normalization gadget, wherein the layer normalization gadget is configured to perform layer normalization operations. Thermodynamic data may be used as input to one or more thermodynamic chips comprising oscillators, wherein thermodynamic evolution according to one or more energy potentials governing the oscillators enable results of layer normalization to be obtained by respective ones of the oscillators. Furthermore, the results may be encoded as thermodynamic data in position degree of freedoms of respective oscillators.
A thermodynamic mixture of experts gadget includes a SoftMax gadget, multiple energy-based models for processing data, and a summation gadget, also called a Selection of Experts gadget. The SoftMax gadget generates one-hot encoded vectors, which correspond to particular ones of the energy-based models for processing data. The outputs of the energy-based models for processing data, in combination with the one-hot encoded vectors, are inputs to the summation gadget, which generates output that is processed data, processed by energy-based models selected by the SoftMax gadget.
A thermodynamic relay gadget includes a relay oscillator and an on-chip controller. The relay oscillator has a time dependent mass or time dependent frequency that is controllable, by the on-chip controller. The relay gadget is configured to relay thermodynamic information in analog form between an output oscillator of a first energy-based model and an input oscillator of a second energy-based model.
G06F 30/28 - Design optimisation, verification or simulation using fluid dynamics, e.g. using Navier-Stokes equations or computational fluid dynamics [CFD]
G01R 33/035 - Measuring direction or magnitude of magnetic fields or magnetic flux using superconductive devices
An analog SoftMax gadget is implemented using one or more thermodynamic chips (neuro-thermodynamic processors). The analog SoftMax gadget takes a thermodynamic input and calculates a result of the SoftMax function thermodynamically according to an engineered potential used for oscillators and oscillator couplings for a set of oscillators that implement the analog SoftMax gadget. The analog SoftMax gadget returns the result of the SoftMax function as a thermodynamic output that may be relayed to other energy-based models of a thermodynamic computer. The input, processing, and output are all performed thermodynamically (e.g., in an analog fashion) without a need to convert the information into a classical representation.
A thermodynamic computing system is implemented using one or more thermodynamic chips that implement a plurality of energy based models (EBMs). Each EBM comprises oscillators, wherein the oscillators represent neuron and synapse values of an engineered energy potential. The synapse values may be updated or trained via mean-field forwards and backwards propagation. During the forwards propagation, the energy potentials of the EBMs are not perturbed, and gradient terms are obtained. During the backwards propagation, the energy potentials of the EBMs are perturbed, and additional gradient terms are obtained. The gradient terms may be combined with the additional gradient terms to determine updated synapse values.
Systems, methods and computer readable media relating to neuro-thermodynamic computers configured to implement one or more components of a transformer neural network architecture, wherein the transformer neural network architecture is configured to perform operations of a transformer neural network. Thermodynamic data may be used as input to one or more thermodynamic chips comprising oscillators, wherein thermodynamic evolution according to one or more energy potentials governing the oscillators enable results of a transformer neural network architecture, or at least intermediate results, to be obtained by respective ones of the oscillators. Furthermore, the results, encoded as thermodynamic data in position degree of freedoms of respective oscillators, of one component may be used as input to another component.
A thermodynamic selection of experts energy-based model gadget includes a SoftMax gadget, a set of oscillators having a potential which is used to modify input data, and multiple energy-based models for processing data. The SoftMax gadget may produce one-hot encoded vectors which may be used by the set of oscillators having a potential which is used to modify input data, such that the modified input data corresponds to one of the multiple energy-based models for processing data.
Systems, methods and computer readable media relating to neuro-thermodynamic computers configured to implement one or more components of a transformer neural network architecture, wherein the transformer neural network architecture is configured to perform operations of a transformer neural network. Thermodynamic data may be used as input to one or more thermodynamic chips comprising oscillators, wherein thermodynamic evolution according to one or more energy potentials governing the oscillators enable results of a transformer neural network architecture, or at least intermediate results, to be obtained by respective ones of the oscillators. Furthermore, the results, encoded as thermodynamic data in position degree of freedoms of respective oscillators, of one component may be used as input to another component.
Systems, methods and computer readable media relating to neuro-thermodynamic computers configured to implement hierarchical architecture, wherein the hierarchical architecture includes one or more layers of components, and wherein the hierarchical architecture is configured to perform Gibbs sampling and nested Gibbs sampling. For example, a block layer may include an energy based model (EBM) implemented using oscillators and couplings between oscillators. A chip layer may include multiple blocks coupled to each other using relay oscillators. A package layer may include multiple chips coupled to each other using additional relay oscillators.
A thermodynamic computing system is implemented using one or more thermodynamic chips that implement a plurality of energy based models (EBMs). Each EBM comprises oscillators, wherein the oscillators represent neuron and synapse values of an engineered energy potential. The synapse values may be updated or trained via mean-field forwards and backwards propagation. During the forwards propagation, the energy potentials of the EBMs are not perturbed, and gradient terms are obtained. During the backwards propagation, the energy potentials of the EBMs are perturbed, and additional gradient terms are obtained. The gradient terms may be combined with the additional gradient terms to determine updated synapse values.
A neuro-thermodynamic computer includes a thermodynamic chip that includes oscillators that are mapped to neurons and additional oscillators that are mapped to synapses, wherein the synapses correspond to weights and bias values used to describe relationships between the neurons in an energy-based model. Learning algorithms are described for computing gradients for a positive phase term and a negative phase term, based on measurements taken of the synapse oscillators, without a need to fully compute updated weights and biases on classical hardware. However, classical hardware may be used to perform basic operations to convert the measured values into calculated updated weights and biases. The updated weights and bias values are used to train the energy-based model, which once trained, can be used to generate inferences for various types of machine learning or AI-type problems.
A neuro-thermodynamic computer includes a thermodynamic chip that includes oscillators that are mapped to neurons and additional oscillators that are mapped to synapses, wherein the synapses correspond to weights and bias values used to describe relationships between the neurons in an energy-based model. Learning algorithms are described for computing gradients for a positive phase term and a negative phase term, as well as elements of an information matrix based on measurements taken of the synapse oscillators, without a need to fully compute updated weights and biases on classical hardware. However, classical hardware may be used to perform basic operations to convert the measured values into calculated updated weights and biases. The updated weights and bias values are used to train the energy-based model, which once trained, can be used to generate inferences for various types of machine learning or AI-type problems.
A neuro-thermodynamic computer includes a thermodynamic processor chip that includes oscillators that are mapped to neurons and additional oscillators that are mapped to synapses, wherein the synapses correspond to weights and bias values used to describe relationships between the neurons in an energy-based model. The neuro-thermodynamic computer further comprises one or more ancilla thermodynamic chips with ancilla oscillators coupled to the synapse oscillators of the thermodynamic processor chip. Learning algorithms are described for computing gradients for a positive phase term and a negative phase term, as well as elements of an information matrix based on measurements taken of the ancilla oscillators, without a need to fully compute updated weights and biases on classical hardware.
A self-learning neuro thermodynamic computing device comprising thermodynamic computing chips as well as systems and methods for performing computing using a self-learning neuro thermodynamic computing device are disclosed. In some embodiments, the self-learning neuro thermodynamic computing device may automatically learn weights and biases to be used for inference generation using Langevin dynamics. In some embodiments, the self-learning neuro thermodynamic computing device comprises two or more coupled thermodynamic chips, such as a clamped thermodynamic chip configured to be clamped to input data (e.g. training data or test data), an un-clamped thermodynamic chip, and a server thermodynamic chip that coordinates between the clamped and un-clamped thermodynamic chips such that weight and bias values are maintained approximately the same between the clamped and un-clamped thermodynamic chips.
09 - Scientific and electric apparatus and instruments
42 - Scientific, technological and industrial services, research and design
Goods & Services
Recorded computer software for signal processing, artificial intelligence (AI) accelerator firmware, artificial intelligence (AI) model generation, artificial intelligence (AI) model generation tools, statistical inference, machine learning, and simulation of stochastic systems; Downloadable computer software for signal processing, artificial intelligence (AI) accelerator firmware, artificial intelligence (AI) model generation, artificial intelligence (AI) model generation tools, statistical inference, machine learning, and simulation of stochastic systems; none of the foregoing in the fields of regulatory compliance, data privacy, data discovery, data remediation, data security, information governance, eDiscovery, digital forensics, or legal investigation Software as a service (SAAS) services featuring software for signal processing, artificial intelligence (AI) accelerator firmware, artificial intelligence (AI) model generation, artificial intelligence (AI) model generation tools, statistical inference, machine learning, and simulation of stochastic systems; Providing on-line non-downloadable software for signal processing, artificial intelligence (AI) accelerator firmware, artificial intelligence (AI) model generation, artificial intelligence (AI) model generation tools, statistical inference, machine learning, and simulation of stochastic systems; none of the foregoing in the fields of regulatory compliance, data privacy, data discovery, data remediation, data security, information governance, eDiscovery, digital forensics, or legal investigation
09 - Scientific and electric apparatus and instruments
35 - Advertising and business services
42 - Scientific, technological and industrial services, research and design
Goods & Services
Computer hardware and software. Provision of an online marketplace for buyers and sellers of
goods and services. Software as a service (SaaS) featuring software.
09 - Scientific and electric apparatus and instruments
35 - Advertising and business services
42 - Scientific, technological and industrial services, research and design
Goods & Services
(1) Computer hardware and software. (1) Provision of an online marketplace for buyers and sellers of goods and services.
(2) Software as a service (SaaS) featuring software.
09 - Scientific and electric apparatus and instruments
42 - Scientific, technological and industrial services, research and design
Goods & Services
Computer hardware; downloadable and recorded computer software for signal processing, artificial intelligence (AI) accelerator firmware, artificial intelligence (AI) model generation, artificial intelligence (AI) model generation tools, statistical inference, machine learning, and simulation of stochastic systems Software as a service (SAAS) featuring software for signal processing, artificial intelligence (AI) accelerator firmware, artificial intelligence (AI) model generation, artificial intelligence (AI) model generation tools, statistical inference, machine learning, and simulation of stochastic systems; providing online, non-downloadable software for signal processing, artificial intelligence (AI) accelerator firmware, artificial intelligence (AI) model generation, artificial intelligence (AI) model generation tools, statistical inference, machine learning, and simulation of stochastic systems
