Get the App
SLTechnology News&Howtos  ›  IT Information  › 

Kunlun Wanwei Tiangong 3.5 model announced that it will launch the invitation test on April 17: realizing intelligent emergence.

Shulou Source: shulou.com Published: 2023-11-24 16:05:28 10月04日 Update

CTOnews.com, April 10 (Xinhua)-- Kunlun Wanwei officially announced that Tiangong 3.5, China's first domestic big language model to truly achieve intelligent emergence, will be released by Kunlun Wanwei and Qidian Zhiyuan, and will launch invitation testing on April 17. In the future, Tiangong 4 and Tiangong 5 will also meet with you.

Emergence phenomenon refers to the complex behavior or characteristics produced in a relatively simple system. In the field of AI, the emergence ability also indicates whether artificial intelligence has a high degree of autonomous learning ability, and whether it is possible to complete complex tasks such as logical reasoning. Recently, many netizens have complained that the competitor model is answering a classic question, "Xiao Wang's father has two sons. The eldest is Da Mao. What's the second name?" Many different answers will be generated according to the wording of the question and the punctuation marks.

Kunlun Wanwei said that the "Tiangong" big model is very close to the intelligence level of OpenAI ChatGPT. ChatGPT is based on the GPT3.5 model, so Kunlun Wanwei named this version "Tiangong" 3.5.

The Kunlun Wanwei R & D team starts from the 100 million level model in 2020, then the billion level model, and then to the 10 billion level model.

* self-developed 10 billion-level model "Yaoguang" data performance in April 2021. In the future, Tiangong 4 and Tiangong 5 will continue to meet with you.

CTOnews.com attached the model of the university of science and technology is about to open an invitation to test address: http://tiangong.kunlun.com

Tags: Model Tiangong Kunlun Wanwei Intelligence different complex ability problem Test two artificial Intelligence Task son team domestic address multiple Learning ability Apple Docker Huawei Linux macOS MariaDB Microsoft MySQL NVidia OPPO Reno MySQL Shulou Technology MariaDB macOS Redmi