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

    Neuronpedia

    INDEX
    Explanations

    hr units and calculations

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     commencent
    0.53
     begin
    0.52
    ENDING
    0.46
     correction
    0.45
    かった
    0.43
     eles
    0.43
    льним
    0.43
     devenu
    0.43
     jedną
    0.43
     elephant
    0.43
    POSITIVE LOGITS
     optimizing
    0.44
     múlti
    0.40
     उत्सर्जन
    0.39
     резо
    0.39
     ორგანო
    0.39
     compactly
    0.39
    dml
    0.39
    Superview
    0.38
    カテゴ
    0.38
     maximizing
    0.38
    Act Density 0.000%

    No Known Activations