Elon Musk unveils Grok-1.5, next-generation AI chatbot with superior coding and math skills
Elon Musk announced that an upgraded version of artificial intelligence company xAI’s chatbot Grok will be released next week.
This revelation came via Musk’s social media. post It was released on March 29, after xAI announced Grok-1.5 in a blog post. The enhanced AI chatbot will initially be accessible to “early testers of social media platforms and existing Grok users.”
Musk also hinted at continued development of Grok 2, which he expects will surpass current AI standards in every respect.
Grock-1.5
Grok-1.5 is an advanced version of the Grok-1 AI model and offers improved inference and a context length of 128,000 tokens.
xAI’s analysis shows that the performance of advanced chatbots has significantly improved, especially for coding and math-related tasks. However, it falls short compared to Google’s Gemini Pro 1.5 and OpenAI’s GPT-4.
According to the company:
“Grok-1.5 achieved a 50.6% score on the MATH benchmark and a 90% score on the GSM8K benchmark. Both math benchmarks cover a wide range of competitive problems from elementary through high school. It also scored 74.1% on the HumanEval benchmark, which evaluates code generation and problem-solving skills.”
Additionally, Grok-1.5 can leverage information from significantly longer documents, and the model can handle longer, more complex prompts while maintaining the ability to follow instructions as the context window expands.
The company added:
“Grok-1.5 is built on a custom distributed training framework based on JAX, Rust, and Kubernetes. This training stack allows our teams to prototype ideas and train new architectures at scale with minimal effort.”
Grok is open source.
Earlier this month, xAI took a major step forward by open sourcing the underlying code for Grok-1.
The decision came in response to legal action filed by Musk against OpenAI, an organization he once co-founded. Musk argued that OpenAI had strayed from its original promise to prioritize open source model development over shareholder interests.
Meanwhile, xAI said the released code is a “raw base model checkpoint from the Grok-1 pre-training phase, which ended in October 2023.” This means that the model is not fine-tuned for specific applications, such as conversations. It added that the model is licensed under the Apache License 2.0.