META RESPONDS TO LLAMA 4 RUMORS

Summary from the TLDR AI newsletter

Meta's VP of Generative AI has denied claims that Llama 4 models were trained on benchmark test sets, pushing back against allegations of inflated performance scores.