A startup claims it broke through a bottleneck that’s holding back LLMs
AI startup Subquadratic claims to have solved a major mathematical bottleneck in large language models with its new 'SubQ' model. While initial skepticism was high due to a lack of data, the company has released third-party benchmarks to support its claims of superior speed and efficiency.
Why it matters
If validated, this technology could significantly reduce the energy and cost requirements for training and running advanced AI models.
Subquadratic has now shared more details about its new model. But some are still skeptical.
The article balances the company's claims with the skepticism of the tech community, maintaining a neutral stance.
Get smarter about the news
Sign up free for a feed built around what you actually care about, Dive Deeper research on any story, and the full text of every article.
Create free accountAlready have an account? Sign in