18097995. VERIFYING REMOTE EXECUTION OF MACHINE LEARNING INFERENCE UNDER HOMOMORPHIC ENCRYPTION USING PERMUTATIONS simplified abstract (International Business Machines Corporation)
Contents
- 1 VERIFYING REMOTE EXECUTION OF MACHINE LEARNING INFERENCE UNDER HOMOMORPHIC ENCRYPTION USING PERMUTATIONS
- 1.1 Organization Name
- 1.2 Inventor(s)
- 1.3 VERIFYING REMOTE EXECUTION OF MACHINE LEARNING INFERENCE UNDER HOMOMORPHIC ENCRYPTION USING PERMUTATIONS - A simplified explanation of the abstract
- 1.4 Simplified Explanation
- 1.5 Key Features and Innovation
- 1.6 Potential Applications
- 1.7 Problems Solved
- 1.8 Benefits
- 1.9 Commercial Applications
- 1.10 Questions about the Technology
- 1.11 Original Abstract Submitted
VERIFYING REMOTE EXECUTION OF MACHINE LEARNING INFERENCE UNDER HOMOMORPHIC ENCRYPTION USING PERMUTATIONS
Organization Name
International Business Machines Corporation
Inventor(s)
Eyal Kushnir of Kfar Vradim (IL)
Ramy Masalha of Kafr Qari (IL)
Nir Drucker of Zichron Yaakov (IL)
VERIFYING REMOTE EXECUTION OF MACHINE LEARNING INFERENCE UNDER HOMOMORPHIC ENCRYPTION USING PERMUTATIONS - A simplified explanation of the abstract
This abstract first appeared for US patent application 18097995 titled 'VERIFYING REMOTE EXECUTION OF MACHINE LEARNING INFERENCE UNDER HOMOMORPHIC ENCRYPTION USING PERMUTATIONS
Simplified Explanation
The patent application describes a technique to remotely identify potential compromise of a service provider that performs homomorphic inferencing on a model. This involves generating permutations of trigger samples and real data samples, encrypting them using a tile tensor data structure, and analyzing the results to determine if the service provider is compromised.
- Homomorphic inferencing on a model performed by a service provider
- Generation of permutations of trigger and real data samples
- Encryption using a tile tensor data structure
- Analysis of results to identify potential compromise of the service provider
Key Features and Innovation
- Utilizes homomorphic inferencing on a model for remote compromise detection
- Generates permutations of trigger and real data samples for analysis
- Encrypts data using a tile tensor data structure for improved performance
- Determines compromise based on the results of the inferencing process
Potential Applications
- Cybersecurity for service providers
- Remote compromise detection in homomorphic inferencing systems
- Data encryption and analysis for sensitive information protection
Problems Solved
- Remote identification of potential compromise in service providers
- Secure analysis of sensitive data samples
- Detection of unauthorized access to homomorphic inferencing models
Benefits
- Enhanced security for service providers
- Improved data privacy and protection
- Efficient detection of compromise in inferencing systems
Commercial Applications
The technology can be applied in industries such as cybersecurity, data analysis, and artificial intelligence. It can be used by service providers to ensure the integrity of their inferencing processes and protect sensitive data from unauthorized access.
Questions about the Technology
What are the potential implications of using homomorphic inferencing for remote compromise detection?
Homomorphic inferencing allows for secure analysis of data without exposing sensitive information, making it a valuable tool for detecting potential compromises in service providers.
How does the encryption using a tile tensor data structure enhance the security of the data samples?
The tile tensor data structure allows for the encryption of multi-dimensional arrays in a secret permutated way, improving the performance and security of the data samples during the inferencing process.
Original Abstract Submitted
A technique to remotely identify potential compromise of a service provider that performs homomorphic inferencing on a model. For a set of real data samples on which the inferencing is to take place, at least first and second permutations of a set of trigger samples are generated. Every set of samples (both trigger and real samples) are then sent for homomorphic inferencing on the model at least twice, and in a secret permutated way. To improve performance, a permutation is packaged with the real data samples prior to encryption using a general purpose data structure, a tile tensor, that allows users to store multi-dimensional arrays (tensors) of arbitrary shapes and sizes. In response to receiving one or more results from the HE-based model inferencing, a determination is made whether the service provider is compromised. Upon a determination that the service provider is compromised, a given mitigation action is taken.
