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

    Neuronpedia

    INDEX
    Explanations

    upgrade to better options

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    ק
    0.78
    ة
    0.70
    די
    0.63
    ز
    0.61
    ين
    0.60
    the
    0.59
    ونة
    0.57
    ية
    0.57
    bahn
    0.56
    きた
    0.55
    POSITIVE LOGITS
     upgrades
    1.12
     upgrade
    1.08
     upgrading
    1.08
     upgraded
    1.00
    升级
    0.95
     to
    0.92
     Upgrade
    0.91
    Upgrade
    0.85
    upgrade
    0.77
     upgrad
    0.75
    Act Density 0.006%

    No Known Activations