Samsung Gauss, generative model for creating texts, code and images

A giant like Samsung certainly couldn’t avoid participating in the competition among the moderns generative models. During the event Samsung AI Forum 2023organized by Samsung Research and focused on generative artificial intelligence, the South Korean company presented Samsung Gauss, underlining how this new product represents a real paradigm shift.

Samsung Gauss is an artificial intelligence model that derives its name from the famous mathematician Carl Friedrich Gauss. There Gaussian distribution or normal distribution theory is a key concept in statistics and mathematics. It is a type of continuous statistical distribution that is symmetrical with respect to its mean and characterized by classical bell shape. It is defined by two main parameters: the mean (or expected value) and the standard deviation.

Company spokespersons explain that the name reflects Samsung’s vision in the field of generative models: the goal is to draw on all phenomena and knowledge to exploit the power of modern artificial intelligence in order to improve the life of every individual, wherever it is.

What the Samsung Gauss generative model is and how it works

Presented with the “unique name” of Samsung Gauss, the generative model developed in the laboratories of the Asian company, however, is made up of three main components: Samsung Gauss Language, Samsung Gauss Code e Samsung Gauss Image.

Samsung Gauss: texts, code, images

Language it is presented as a language model designed to improve productivity by helping with tasks such as writing emails, summarizing texts and documents, processing translations. It can also improve the user experience by providing more immediate and intelligent control of each device when applied to individual products. The model itself is made up of various components which make it easily adaptable for use in localeon a specific device, or on the cloud.

Code instead, it is a coding assistant that helps programmers develop working algorithms and applications in a short time, especially in corporate and team environments. The system presented by Samsung is equipped with an interactive interface that supports functions such as generating comments for various portions of code, creating tests to verify functionality, automatic completion and much more.

Image It also allows you to generate and edit images. The model can not only generate new works but also modify existing photos and images by making optimizations, adding or removing objects, converting low resolution images into high resolution (upscaling).

Samsung Glass will be integrated into the company’s products

For the moment, Samsung Glass is widely used by the employees and collaborators of the South Korean giant. The company specifies, however, that generative models will soon also land on devices designed, manufactured and marketed by Samsung.

Although there is no official confirmation yet, sources close to the company reveal that the first products to integrate Samsung Glass will be smartphone Galaxy S24which is scheduled for release in January 2024.

On the other hand, Samsung brand devices (as well as those of other manufacturers) increasingly focus on NPU (Neural Processing Unit), processing units specifically designed to accelerate and manage calculations related to artificial neural networks, the machine learning and artificial intelligence applications. Since Samsung Glass can also work locally, it is reasonable to expect that generative models will take full advantage of future, increasingly powerful NPUs.

Samsung has also established a AI Red Team, to verify the application of ethical principles in the use of artificial intelligence solutions. This specialized team internally monitors privacy and security issues during the various phases of creating, managing, updating and deploying models: from data collection to results formation.

LEAVE A REPLY

Please enter your comment!
Please enter your name here