PANews reported on March 29 that according to Techcrunch, Musk’s artificial intelligence startup X.ai announced its latest generative artificial intelligence model Grok-1.5. The model will power social network X’s Grok chatbot in the near future (“in the coming days,” X.ai wrote in a blog post). Judging from the benchmark results and specifications released by X.ai, the Grok-1.5 appears to be an upgrade from its predecessor, the Grok-1.
According to X.ai, Grok-1.5 has improved in terms of "reasoning capabilities," especially in programming and math-related tasks. The model outperformed Grok-1 by more than twice the score on the popular mathematics benchmark MATH, while improving scores by more than 10% on the HumanEval programming language generation and problem-solving ability tests. A clear improvement of Grok-1.5 over Grok-1 is the amount of textual context it can handle. Grok-1.5 has a context of 128,000 Tokens - "Tokens" referring to fragments of original text. X.ai writes in the aforementioned blog post: “Grok-1.5 can leverage information from longer documents. Additionally, as its context window expands, the model can handle longer, more complex documents while maintaining its ability to follow instructions. Complex prompts.”
X.ai says Grok-1.5 will soon be available to early testers of X and will be accompanied by "several new features."
