Patents/Inventors

Marc Tremblay

Bellevue, WAUS22 patents

21 Patents

US125859342026
Compressing Tokens Based on Positions for Transformer Models
Microsoft Technology Licensing, LLC
0 cites
US124120352025
Token Packing for Sequence Models
Microsoft Technology Licensing, LLC
0 cites
US123803322025
Forcing Weights of Transformer Model Layers
Microsoft Technology Licensing, LLC
0 cites
US123539842025
Hardware-assisted Gradient Optimization Using Streamed Gradients
Microsoft Technology Licensing, LLC
0 cites
US121827162024
Compressing and Decompressing Data for Language Models
Microsoft Technology Licensing, LLC
0 cites
US121504172024
Systems and Methods for Managing Undesired Effects Produced by Horticultural Lighting Sources in an Electrical Grid
SOLLUM TECHNOLOGIES Inc.
0 cites
US120823202024
Systems and Methods for Monitoring and Managing Reactive Power from Horticultural Lighting Sources in an Electrical Grid
SOLLUM TECHNOLOGIES Inc.
0 cites
US120618482024
Method and System for Simulating and Implementing Horticultural Lighting Scenarios
Sollum Technologies Inc.
0 cites
US119544482024
Determining Position Values for Transformer Models
Microsoft Technology Licensing, LLC
0 cites
US119284292024
Token Packing for Sequence Models
Microsoft Technology Licensing, LLC
0 cites
US118934692024
Position Masking for Transformer Models
Microsoft Technology Licensing, LLC
0 cites
US118869832024
Reducing Hardware Resource Utilization for Residual Neural Networks
Microsoft Technology Licensing, LLC
0 cites
US117068602023
Horticultural Lighting Scenarios Management System and Method
SOLLUM TECHNOLOGIES Inc.
0 cites
US116819052023
Hardware-assisted Gradient Optimization Using Streamed Gradients
Microsoft Technology Licensing, LLC
0 cites
US116756542023
Systems and Methods for Error Recovery
Microsoft Technology Licensing, LLC
0 cites
US116634442023
Pipelined Neural Network Processing with Continuous and Asynchronous Updates
Microsoft Technology Licensing, LLC
0 cites
US116512282023
Dual-momentum Gradient Optimization with Reduced Memory Requirements
Microsoft Technology Licensing, LLC
0 cites
US116153012023
Lossless Exponent and Lossy Mantissa Weight Compression for Training Deep Neural Networks
MICROSOFT TECHNOLOGY LICENSING, LLC
0 cites
US116101202023
Systems and Methods for Training a Neural Network
Microsoft Technology Licensing, LLC
0 cites
US115620472023
Accelerator for Dense and Sparse Matrix Computations
Microsoft Technology Licensing, LLC
0 cites
US115445372023
Token-position Handling for Sequence Based Neural Networks
Microsoft Technology Licensing, LLC
0 cites