By Gleb Bryanski and Elena Fabrichnaya ST PETERSBURG (Reuters) -Russia’s largest lender, Sberbank, plans to unveil a version of its Gigachat large language model (LLM) with reasoning capabilities, First Deputy CEO Alexander Vedyakhin told Reuters in an interview. “In the near future, we will release a new model. Our GigaChat will have reasoning functions, meaning […]
Science
Russia’s Sberbank plans to unveil LLM with reasoning capacity
Audio By Carbonatix
By Gleb Bryanski and Elena Fabrichnaya
ST PETERSBURG (Reuters) -Russia’s largest lender, Sberbank, plans to unveil a version of its Gigachat large language model (LLM) with reasoning capabilities, First Deputy CEO Alexander Vedyakhin told Reuters in an interview.
“In the near future, we will release a new model. Our GigaChat will have reasoning functions, meaning it will be capable of scientific research,” Vedyakhin said. “I am already testing the beta version.”
Such models are capable of reasoning through complex tasks and can solve more challenging problems than previous models in areas such as science, coding, and math. Global market leader OpenAI launched its reasoning LLM last September.
Vedyakhin said Russia still lags behind AI leaders the United States and China by six to nine months, but language adaptation, stable performance, and the use of domestic cloud servers make Sberbank’s GigaChat a preferred model for Russian corporate clients.
Vedyakhin said that around 15,000 companies in Russia are already using Sberbank’s GigaChat. Yandex, Sberbank’s main domestic rival in artificial intelligence, said in May that its search engine now had reasoning functionality.
(Reporting by Elena Fabrichnaya and Gleb Bryanski; Editing by Mark Trevelyan)