The new AI model, introduced in August, uses a similar “train of thought” to the o1 model recently released by OpenAI, to make reliable judgments.
Accordingly, the principle of the technique is to break down complex problems into smaller logical steps. Therefore, it helps to improve the accuracy of responses in difficult fields such as science, programming and mathematics.
Currently, AI models still need to undergo reinforcement learning through human feedback, a process that is expensive, inefficient, and often requires the involvement of experts to accurately label data or verify answers to highly demanding queries.
Currently, Meta is still the only company releasing these models to the public.
Additionally, Meta also released an update to its Segment Anything image recognition model, which speeds up the generation of data that can be used in new inorganic materials research.
Source: https://kinhtedothi.vn/meta-ra-mat-mo-hinh-ai-moi.html
Comment (0)