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

    Neuronpedia

    INDEX
    Explanations

    code closing brace and console.log

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     But
    0.86
     ف
    0.82
     Hence
    0.79
     So
    0.79
     という
    0.78
     Beneath
    0.77
     Під
    0.77
     فَ
    0.77
     ول
    0.76
    américa
    0.75
    POSITIVE LOGITS
    .,
    0.79
    Approach
    0.74
    approach
    0.73
    appro
    0.67
     판
    0.65
     approach
    0.64
     olla
    0.63
     hab
    0.63
    wärm
    0.62
     approaches
    0.62
    Act Density 0.060%

    No Known Activations