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

    Neuronpedia

    INDEX
    Explanations

    specific actions

    np_max-act-logits · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    我已经
    0.69
    nobody
    0.59
     фору
    0.57
     dungeons
    0.57
    forums
    0.56
     wären
    0.54
    都已经
    0.54
     امکان
    0.53
     deja
    0.52
     Lockdown
    0.52
    POSITIVE LOGITS
    Percent
    0.68
     creatives
    0.64
     sorprendente
    0.62
     surprising
    0.61
     উদার
    0.59
    糁
    0.59
     broader
    0.58
    යා
    0.58
     sorprend
    0.58
    弹性
    0.58
    Act Density 0.001%

    No Known Activations