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

    Neuronpedia

    INDEX
    Explanations

    add honey

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     hitherto
    0.71
     thence
    0.64
     sobretudo
    0.60
     putative
    0.56
     consequent
    0.55
     impossibility
    0.55
     nigga
    0.55
     instructive
    0.54
     variously
    0.54
    itherto
    0.53
    POSITIVE LOGITS
     honey
    0.69
    Honey
    0.66
    honey
    0.62
    蜂蜜
    0.58
     Honey
    0.57
    很多的
    0.55
    🍯
    0.52
    蜜
    0.51
     sweeter
    0.50
     ولكن
    0.48
    Act Density 0.016%

    No Known Activations