Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    INDEX
    Explanations

    available to the public

    np_acts-logits-general · gemini-2.5-flash-lite

    mentions that the model is an open-weights (open-source) model widely available to the public.

    oai_token-act-pair · gpt-5-miniTriggered by @yooniel31
    New Auto-Interp
    Negative Logits
     dimers
    0.43
     catt
    0.43
     চাওয়া
    0.41
     compounds
    0.41
     stumpage
    0.41
    Compounds
    0.41
     adjournment
    0.40
     clos
    0.40
     ہو
    0.40
     stratification
    0.40
    POSITIVE LOGITS
     контроли
    0.41
    fitrión
    0.40
     responsibly
    0.40
     unlike
    0.39
    प्रयोग
    0.39
    unlike
    0.39
     externos
    0.39
    free
    0.39
    ემის
    0.38
    open
    0.38
    Act Density 0.383%

    No Known Activations