Llama 4 Scout and Maverick Face Scrutiny Over Benchmarking Practices. Meta released two new open-weight large language models, dubbed Llama 4 Scout and Llama 4 Maverick, that aim to scale AI performance while lowering compute costs. Their debut drew criticism over how the company presented benchmark results.
First seen on govinfosecurity.com
Jump to article: www.govinfosecurity.com/metas-new-llama-4-models-stir-controversy-a-27949
