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

    Neuronpedia

    INDEX
    Explanations

    contractions and possessive forms in the text

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

    or numbers after apostrophes

    np_acts-logits-general · gemini-2.0-flash

    possessive 's

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     öv
    -0.54
     själva
    -0.54
    ReusableCell
    -0.53
    IBOutlet
    -0.52
    NSUserDefaults
    -0.51
    liament
    -0.51
    vedať
    -0.49
    encils
    -0.47
    Чыганаклар
    -0.47
    المناصب
    -0.47
    POSITIVE LOGITS
     himself
    0.82
     الحره
    0.69
    BeginContext
    0.66
     ujednoznacz
    0.65
     <=",
    0.65
     husband
    0.61
     stepfather
    0.60
     him
    0.58
     Himself
    0.57
     memu
    0.57
    Act Density 0.163%

    No Known Activations