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

    Neuronpedia

    INDEX
    Explanations

    is checks states or properties

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     Find
    0.84
    いま
    0.79
     crashes
    0.79
    skills
    0.78
     gap
    0.78
    Find
    0.77
    有無
    0.75
    чнее
    0.75
    oną
    0.73
     holes
    0.71
    POSITIVE LOGITS
     זה
    1.18
    abel
    1.07
     féidir
    1.00
    uzu
    1.00
    abella
    0.95
    լ
    0.95
     뭔
    0.92
     الموس
    0.92
    uš
    0.92
     setIs
    0.92
    Act Density 0.200%

    No Known Activations