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

    Neuronpedia

    INDEX
    Explanations

    scaling between, truly becoming

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     diced
    0.68
     Hide
    0.65
    0.65
     North
    0.64
     Commit
    0.63
     TAP
    0.62
     Residences
    0.62
     Dirty
    0.62
     Jones
    0.61
     Recycle
    0.61
    POSITIVE LOGITS
     zemlji
    0.70
    ulp
    0.63
    zących
    0.63
    gerät
    0.63
    Gew
    0.62
    Synth
    0.61
    गोवा
    0.60
    मंगल
    0.60
    Wel
    0.60
    Overall
    0.60
    Act Density 0.000%

    No Known Activations