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

    Neuronpedia

    INDEX
    Explanations

    suggesting options and examples

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     rohkem
    0.48
     manejar
    0.39
     bullshit
    0.38
     StringSet
    0.38
     be
    0.37
     mettre
    0.37
     die
    0.36
     unobstructed
    0.36
     interfere
    0.36
    ك
    0.36
    POSITIVE LOGITS
    甚至
    0.59
     मसलन
    0.54
     thậm
    0.51
     навіть
    0.51
    например
    0.49
    even
    0.49
     기준으로
    0.48
     เช่น
    0.48
     даже
    0.48
    甚至是
    0.48
    Act Density 0.022%

    No Known Activations