X (formerly Twitter) has announced the launch of the new version of its ‘chatbot’, Grok-1.5, which comes with improvements in your ability to perform on tasks related to coding and mathematics.
The company presented this generative Artificial Intelligence (AI) model in November of last year, when it noted that it was capable of offering responses in real time with a humor that is not suitable for all users.
It is a ‘chatbot’ inspired by one of the favorite novels of the owner of the social network, Elon Musk, and initially was aimed at Premium+ users, although its availability has been expanded to the rest of Premium subscribers this week.
The company has now announced Grok-1.5, the new iteration of its ‘chatbot’, which comes with improved reasoning capabilities on tasks related to coding and mathematics.
X has indicated that, during its performance tests, Grok-1.5 achieved a 50.6 percent score in the comparative test of mathematics and a 90 percent score in which it corresponds to GSM8K.
The company has specified that these two mathematics benchmark tests “cover a wide range of competency problems from elementary school to high school,” according to a statement.
In this he has also said that Grok-1.5 scored 74.1 percent on the HumanEval benchmark, which evaluates code generation and problem-solving capabilities.
Furthermore, X has pointed out that a new feature of this iteration is the ability to process long contexts of up to 128,000 tokens within its context window, allowing the ‘chatbot’ to have an increased memory capacity of up to 16 times the length of the previous context. This way you can use information from “substantially longer” documents.
The firm has also said that the model can handle longer and more complex prompts “while maintaining the ability to follow instructions as they occur.” its contextual window expands”.
Grok-1.5 “should be available next week” according to Elon Musk for some users, from whom X hopes to receive comments to improve the tool. Additionally, he hopes to introduce “several new features” in the coming days as he brings it to a broader audience.