LG open source EXAONE Deep inference AI model, the first self-developed inference AI model in South Korea

LG AI Research recently opened the globally esteemed EXAONE Deep inference AI model. The biggest highlight of this model lies in its inference ability. EXAONE Deep is able to independently formulate assumptions and conduct inference verification to make independent decisions, which marks a new era of active AI. LG AI Research is proud to announce that EXAONE Deep is the first self-developed inference AI model in South Korea to achieve this milestone.

EXAONE Deep-32B has 32 billion parameters, and although it is only 5% of some giant models, it has shown amazing strength in solving complex problems. Performance evaluation results released on Hugging Face, a global open source AI platform, show that EXAONE Deep-32B has performed outstandingly in logical reasoning and problem-solving capabilities, especially in the field of mathematics in the Korean college entrance examination. It scored the highest score of 94.5 and reached the highest level in all elective subjects.

In addition to the powerful 32B version, LG AI Research also open sourced the lightweight model EXAONE Deep-7.8B and the end-side model EXAONE Deep-2.4B simultaneously. The performance of these two models is still remarkable when the size is greatly reduced. The end-side model has unique advantages in terms of security and can better protect users' security and personal information. LG AI Research expects that this end-side AI model will be rapidly popularized in various industries such as smartphones, cars and robots.

Project entrance: https://top.aibase.com/tool/exaone-deep