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

    Neuronpedia

    INDEX
    Explanations

    dismantle corrupt entities

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    ↵↵
    0.91
    ↵
    0.75
    user
    0.68
    name
    0.68
    2
    0.64
    .
    0.63
    concept
    0.59
    but
    0.58
     y
    0.58
    cur
    0.57
    POSITIVE LOGITS
    <unused2150>
    1.20
    <unused1150>
    1.10
    <unused1040>
    1.09
    <unused2078>
    1.09
     newApproved
    1.07
    <unused241>
    1.07
     Polynucleaires
    1.06
    <unused91>
    1.06
    ecocond
    1.06
    <unused618>
    1.05
    Act Density 0.013%

    No Known Activations