1 views collected around this technical thread.
Meta AI Vice President Ahmad Al‑Dahle addressed recent criticisms of the newly released Llama 4 model, denying claims of test‑set cheating, explaining quality variations as post‑release optimization, and acknowledging internal concerns that led to staff resignations and calls for transparency.
The article introduces Meta's newly open‑sourced Llama 4 series—including Scout with a 1 billion‑token context window, Maverick with 400 billion parameters, and the upcoming Behemoth teacher model—detailing their expert‑mix architecture, the NoPE positional‑encoding removal, training pipelines, performance benchmarks, and infrastructure improvements for large‑scale AI research.
Meta's newly released Llama 4 quickly became a controversy as internal leaks reveal training‑data cheating, benchmark over‑optimization, and disappointing code‑generation performance that fails to match even older models, prompting resignations and widespread criticism from the AI community.
OpenAI announced that GPT‑5 will be delayed yet offered for free, with upcoming releases of o3 and o4‑mini, while also promising an open‑source inference model, whereas Meta’s Llama 4 faces repeated postponements amid performance concerns and a massive AI infrastructure investment.
Meta announced the Llama 4 series—Scout, Maverick and Behemoth—featuring multimodal capabilities, Mixture‑of‑Experts design, up to 10 million‑token context windows, and state‑of‑the‑art performance on STEM, multilingual and image benchmarks, with models now downloadable from llama.com and Hugging Face.