IP Library Patent Application 19088978
Patent Application
App. No. 19/088,978

DEEP LEARNING USING LARGE CODEWORD MODEL WITH HOMOMORPHICALLY COMPRESSED DATA

Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US None
App. No.
19/088,978
Filed
Mar 24, 2025
Art Unit
2845
USPC
707/693
Abstract

A system and method for deep learning using a large codeword model with homomorphically compressed and dyadically encrypted data is disclosed. The system preprocesses input data, applies homomorphic-dyadic compression and encryption, tokenizes the compressed data into sourceblocks, and assigns codewords using a codebook. These codewords are processed through a machine learning core, which can be either a conventional transformer-based architecture or a latent transformer core utilizing a variational autoencoder. The system enables secure operations on encrypted data, preserving privacy while allowing complex computations. The processed output is decrypted, decompressed, and translated to match the input modality. A neural upsampler may further enhance the output. The machine learning core is continuously trained using the processed data and additional training data, improving performance over time.

Claims (65)

1 . A computer system comprising:

a hardware memory, wherein the computer system is configured to execute software instructions stored on nontransitory machine-readable storage media that:

compress a plurality of input data sets by:

analyzing the input data sets to determine statistical properties and frequency distributions;

creating one or more transformation matrices based on the determined properties, wherein the transformation matrices are configured to preserve homomorphic properties of the input data sets;

transforming the input data sets using the transformation matrices to generate modified probability distributions;

generating (i) main data streams comprising the transformed data and (ii) secondary data streams comprising transformation information; and

compressing the main data streams while maintaining the homomorphic properties;

tokenize the compressed main data streams into a plurality of sourceblocks;

assign the plurality of sourceblocks a plurality of codewords, where each sourceblock is mapped to a particular codeword through a codebook;

process the plurality of codewords through a machine learning core to generate a codeword response;

translate the codeword response into a translated response which matches the modality of the inputs; and

decompress and decrypt the translated response.

2 . The system of claim 1 , wherein the machine learning core is a conventional transformer-based architecture comprising:

an embedding layer;

a positional encoding layer; and

a series of transformer layers.

3 . The system of claim 2 , further comprising a syntactic splitting component that splits the codewords into smaller units before processing through the conventional transformer-based architecture.

4 . The system of claim 1 , wherein the machine learning core is a latent transformer core comprising:

a variational autoencoder with an encoder and a decoder; and

a transformer that processes latent space vectors, wherein the transformer does not include an embedding layer and a positional encoding layer.

5 . The system of claim 4 , wherein processing the plurality of codewords through the machine learning core comprises:

generating a plurality of latent space vectors by processing the plurality of codewords through the variational autoencoder's encoder;

learning relationships between the plurality of latent space vectors by processing them through the transformer;

using the learned relationships to generate a plurality of output latent space vectors; and

generating the codeword response by passing the output latent space vectors through the variational autoencoder's decoder.

6 . The system of claim 5 , further comprising a syntactic splitting component that splits the latent space vectors into smaller units before processing through the transformer.

7 . The system of claim 1 , wherein compressing and encrypting the input data sets further comprises:

combining the compressed main data streams and the secondary data streams into output streams; and

implementing security measures to protect the output streams.

8 . The system of claim 7 , wherein the security measures comprise providing cryptographically secure random numbers for use in data transformation and implementing protections against side-channel attacks.

9 . The system of claim 1 , wherein transforming the input data into modified distributions comprises transforming the input data into dyadic distributions.

10 . The system of claim 1 , further comprising a neural upsampler that processes the codeword response to generate a reconstructed output containing more information than the translated response.

11 . A method for deep learning using a large codeword model with homomorphically compressed dyadically encrypted data, comprising the steps of:

compressing a plurality of input data sets by:

analyzing the input data sets to determine statistical properties and frequency distributions;

creating one or more transformation matrices based on the determined properties, wherein the transformation matrices are configured to preserve homomorphic properties of the input data sets;

transforming the input data sets using the transformation matrices to generate modified probability distributions;

generating (i) main data streams comprising the transformed data and (ii) secondary data streams comprising transformation information; and

compressing the main data streams while maintaining the homomorphic properties;

tokenizing the compressed main data streams into a plurality of sourceblocks;

assigning the plurality of sourceblocks a plurality of codewords, where each sourceblock is mapped to a particular codeword through a codebook;

processing the plurality of codewords through a machine learning core to generate a codeword response;

translating the codeword response into a translated response which matches the modality of the inputs; and

decompressing and decrypt the translated response.

12 . The method of claim 11 , wherein the machine learning core is a conventional transformer-based architecture comprising:

an embedding layer;

a positional encoding layer; and

a series of transformer layers.

13 . The method of claim 12 , further comprising splitting the codewords into smaller units before processing through the conventional transformer-based architecture.

14 . The method of claim 11 , wherein the machine learning core is a latent transformer core comprising:

a variational autoencoder with an encoder and a decoder; and

a transformer that processes latent space vectors, wherein the transformer does not include an embedding layer and a positional encoding layer.

15 . The method of claim 14 , wherein processing the plurality of codewords through the machine learning core comprises:

generating a plurality of latent space vectors by processing the plurality of codewords through the variational autoencoder's encoder;

learning relationships between the plurality of latent space vectors by processing them through the transformer;

using the learned relationships to generate a plurality of output latent space vectors; and

generating the codeword response by passing the output latent space vectors through the variational autoencoder's decoder.

16 . The method of claim 15 , further comprising splitting the latent space vectors into smaller units before processing through the transformer.

17 . The method of claim 11 , wherein compressing and encrypting the input data sets further comprises:

combining the compressed main data streams and the secondary data streams into output streams; and

implementing security measures to protect the output streams.

18 . The method of claim 17 , wherein the security measures comprise providing cryptographically secure random numbers for use in data transformation and implementing protections against side-channel attacks.

19 . The method of claim 11 , wherein transforming the input data into modified distributions comprises transforming the input data into dyadic distributions.

20 . The method of claim 11 , further comprising processing the codeword response through a neural upsampler to generate a reconstructed output containing more information than the translated response.

Assignments (1)
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Apr 7, 2025
From: GALVIN, BRIAN
To: ATOMBEAM TECHNOLOGIES INC.
Reel/Frame 070759/0034 →