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

    Neuronpedia

    INDEX
    Explanations

    phrases indicating the presence of exceptions or singular cases within a context

    oai_token-act-pair · gpt-4o-miniTriggered by @bot

    all but one, with one exception

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     contextLoads
    -0.32
     casada
    -0.30
     participaron
    -0.30
    ună
    -0.30
    ⟬
    -0.28
     povezave
    -0.28
    feitura
    -0.28
    ografija
    -0.27
    encodeWith
    -0.27
    چه
    -0.27
    POSITIVE LOGITS
     exceptions
    0.82
     exception
    0.79
     Exceptions
    0.76
     protoimpl
    0.75
     Exception
    0.70
     excep
    0.68
    exceptions
    0.65
    findpost
    0.65
    Except
    0.65
    例外
    0.65
    Act Density 0.062%

    No Known Activations