Meta Steps Into The Artificial Intelligence Paint With Llama 3: Is It A Top 3 Bot?

Meta, the tech giant formerly known as Facebook, is making waves in the artificial intelligence (AI) arena with its latest creation, Llama 3. Touted as a significant leap forward in AI Technology, Llama 3 boasts it can outperform most current AI models, including Google’s Gemini.

In a recent blog post, Meta highlighted Llama 3’s superiority over existing models, citing improvements in diversity of responses, reduced false refusals, and enhanced reasoning capabilities. Llama 3 has two model weights available.

“This evaluation set contains 1,800 prompts that cover 12 key use cases: asking for advice, brainstorming, classification, closed question answering, coding, creative writing, extraction, inhabiting a character/persona, open question answering, reasoning, rewriting, and suMMArization,” Meta says in the blog post, The Verge reported.

Meta’s benchmark tests reveal that Llama 3 surpasses similarly sized models such as Google’s GeMMA and Gemini, as well as Mistral 7B and Anthropic’s Claude 3. In particular, Llama 3 8B excelled in the MMLU benchmark, showcasing its superiority in general knowledge, The Verge reported.

Human evaluators also ranked Llama 3 higher than other models, including OpenAI’s GPT-3.5. Meta created a new dataset to emulate real-world scenarios, allowing for comprehensive evaluation across various use cases, from advice-seeking to creative writing.

Meta plans to expand Llama 3’s capabilities with larger model sizes, enabling more complex responses such as image generation and audio transcription.

Yann LeCun, Meta’s chief AI scientist, hinted at an even more powerful successor to Llama 3, poised to rival top closed AI models like GPT-4 and Google’s Gemini. LeCun’s optimism about the potential of open-source AI models to accelerate progress towards human-level intelligence signals a promising future for AI innovation.

“It takes a lot of time to fine-tune, but a bunch of variations of these models are going to come out in the next few months,” LeCun said at Imagination in Action, a conference at MIT focused on generative AI, Wired reported.

He also said, “AI is better when more people look at the code. Infrastructure needs to be open source—it just progresses faster.”

Meta logo, Wiki Commons/Photo by Dmitry Zvolskiy: https://www.pexels.com/photo/white-and-brown-llama-2253639/