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

    Neuronpedia

    INDEX
    Explanations

    versions comparing latest and older

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    ган
    0.67
     pretext
    0.67
    ウンド
    0.66
    ocou
    0.64
    tres
    0.63
     Bid
    0.60
    userdata
    0.60
     intrac
    0.59
     left
    0.58
     HUD
    0.58
    POSITIVE LOGITS
    Version
    1.74
    version
    1.73
     Version
    1.69
     version
    1.61
    バージョン
    1.60
     Versions
    1.60
    versions
    1.60
     versions
    1.57
    版本
    1.57
     版本
    1.56
    Act Density 0.619%

    No Known Activations