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

    Neuronpedia

    INDEX
    Explanations

    non-latin characters

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     inclined
    0.49
     compressing
    0.46
     didnt
    0.46
     invece
    0.46
     instead
    0.45
     died
    0.45
     evicted
    0.45
     illumination
    0.44
     ervoor
    0.44
     entertained
    0.44
    POSITIVE LOGITS
    灵
    0.43
    CHEMY
    0.43
    ACS
    0.41
    Beam
    0.41
    手
    0.41
    周期
    0.40
     наход
    0.40
    தான
    0.40
    ви
    0.39
     мето
    0.39
    Act Density 0.015%

    No Known Activations