The most recent generative AI model developed by Elon Musk’s AI company, X.ai, is called Grok-1.5. According to a blog post, Grok-1.5 will power the Grok chatbot on social network X “in the coming days.” Based on benchmark results and specifications that have been made public, Grok-1.5 looks to be a measurable improvement over Grok-1.
X.ai claims that Grok-1.5 benefits from “improved reasoning,” especially when it comes to coding and math-related tasks. The model beat Grok-1 by almost two times on the widely used math benchmark, MATH. It also outperformed Grok-1 by more than ten percentage points on the HumanEval test, which measures programming language generation and problem-solving skills.
It’s challenging to predict how such findings will come out in real-world usage. Commonly used AI requirements, which evaluate performance on esoteric tasks like graduate-level chemistry test questions, do not really represent how ordinary people interacts with models in the real world today, as we recently noted.
The amount of context that Grok-1.5 can understand in comparison to Grok-1 is one improvement that should result in noticeable advantages.
Up to 128,000 token contexts can be processed by Grok-1.5. Here, “tokens” refers to unprocessed fragments of text (for example, the word “fantastic” split into “fan,” “tas,” and “tic”). Context, also known as the context window, is the set of input data (text in this case) that a model takes into consideration before producing output (additional text). Larger contexts help models avoid this problem and, as an added benefit, help them understand the flow of data they get. Models with limited context windows have an ability to forget the contents of even recent discussions.
According to X.ai’s blog post, “[Grok-1.5 can] utilize information from substantially longer documents.” Additionally, when the context window develops, the model can respond to longer and more complicated suggestions while still being able to follow instructions.
X.ai’s Grok models have separated themselves from other generative AI models in the past by answering queries regarding subjects such as conspiracies and more controversial political concepts that are normally off-limits to other models. Additionally, Musk has described the models’ responses to inquiries as having “a rebellious streak” and using openly offensive words when asked.
What, if any, modifications Grok-1.5 makes in these areas is unknown. This is not mentioned by X.ai in the blog post.
Early testers on X will soon have access to Grok-1.5 along with “several new features.” Musk has already made references to collecting conversations and responses and offering ideas for posts; we’ll see if things materialize quickly enough.
The news follows X.ai’s open-sourcing of Grok-1, despite the reality that it lacked the programming needed to improve or further train it. More recently, Musk announced that the Grok chatbot, which was previously exclusive to X Premium+ customers (who pay $16 per month), will be accessible to more X users, particularly those who were paying for X’s $8 per month Premium plan.
Topics #AI #Artificial intelligence #Chatbot #Elon Musk #Grok #Grok AI #Grok AI Chabot #news #Twitter #X
