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

    Neuronpedia

    INDEX
    Explanations

    the word "step" or related terms like "feature"

    oai_token-act-pair · gemini-2.0-flash

    step

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Negative Logits
     feroit
    -0.79
     sauvages
    -0.78
     fédé
    -0.74
     mourut
    -0.74
     avoient
    -0.74
     réfrig
    -0.73
     religieuses
    -0.73
     myö
    -0.72
     aveug
    -0.72
     étoient
    -0.72
    POSITIVE LOGITS
    idal
    0.71
     Empieza
    0.68
    estre
    0.68
     getItemId
    0.67
    Collegamenti
    0.67
     thức
    0.67
     NgModule
    0.67
    allar
    0.66
    pade
    0.66
     mar
    0.65
    Act Density 0.953%

    No Known Activations