Meta exec denies the company artificially boosted Llama 4’s benchmark scores

  • staffstaff
  • AI
  • April 7, 2025
  • 0 Comments

A Meta exec on Monday denied a rumor that the company trained its new AI models to present well on specific benchmarks while concealing the models’ weaknesses. The executive, Ahmad Al-Dahle, VP of generative AI at Meta, said in a post on X that it’s “simply not true” that Meta trained its Llama 4 Maverick […]

  • Related Posts

    Ex-Meta engineer raises $14M for Lace AI, a revenue generation software startup

    As an AI engineer at Meta, Boris Valkov helped build PyTorch, one of the world’s largest machine learning libraries. During his time there, Valkov realized that artificial intelligence “was about…

    Continue reading
    Two undergrads built an AI speech model to rival NotebookLM

    A pair of undergrads, neither with extensive AI expertise, say that they’ve created an openly available AI model that can generate podcast-style clips similar to Google’s NotebookLM. The market for…

    Continue reading

    Leave a Reply

    Your email address will not be published. Required fields are marked *

    You Missed

    Ex-Meta engineer raises $14M for Lace AI, a revenue generation software startup

    • By staff
    • April 22, 2025
    • 0 views

    Adaptive Computer wants to reinvent the PC with ‘vibe’ coding for non-programmers

    • By staff
    • April 22, 2025
    • 1 views

    Two undergrads built an AI speech model to rival NotebookLM

    • By staff
    • April 22, 2025
    • 1 views

    Crowdsourced AI benchmarks have serious flaws, some experts say

    • By staff
    • April 22, 2025
    • 3 views