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

    Neuronpedia

    INDEX
    Explanations

    scarecrow

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Negative Logits
     detay
    -0.07
    .tw
    -0.07
     φω
    -0.06
     Raid
    -0.06
    .problem
    -0.06
     bò
    -0.06
     pearl
    -0.06
     光
    -0.06
     kdo
    -0.05
    -example
    -0.05
    POSITIVE LOGITS
     �
    0.08
    ユ
    0.07
     keeping
    0.07
    مم
    0.07
     underscore
    0.07
     principales
    0.07
     ближ
    0.07
    _SLAVE
    0.07
    žel
    0.07
     Braves
    0.07
    Act Density 0.001%

    No Known Activations