Meta Releases MobileLLM – R1 Series of Small Models with Leading Performance in Coding Tests - AI

7 x 24 Track global technological trends

Hot Topic

Day

News Topic

Meta Releases MobileLLM – R1 Series of Small Models with Leading Performance in Coding Tests

2 day ago / Read about 0 minute

Author：小编

Meta's MobileLLM-R1 series of small inference models adopts a 'depth-narrowing' architectural design, offering three parameter sizes: 140 million, 360 million, and 950 million. These models focus on mathematical reasoning, code generation, and scientific computing, demonstrating leading performance in the LiveCodeBench coding test, making them highly suitable as localized code assistance tools for development environments.

Previous page：AI Emerges as a New Growth Engine, with Tencent Pl...

Next page：Mistral AI Unveils New Open-Source Reasoning Model...

Return to List

Hot Reading

2 day ago

Nvidia CEO Jensen Huang says that DGX Spark is powered by N1, confirming N1 SoC and GB10 Superchip as the same

2 day ago

TEAC's new CD deck promises a "pristine" signal – just don't look at the price

2 day ago

Gemini Now Available for Free on Chrome, But There's a Catch for a Subscription-Free AI Experience

2 day ago

Engineering the Future of Databases: How Ganesh Nerella Is Defining Cloud Transformation in the AI Era