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

    Neuronpedia

    INDEX
    Explanations

    breakdown of

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     тыс
    1.02
    <unused550>
    1.02
     ẞ
    1.01
    第壹百
    1.01
    ️
    1.00
    𒊬
    0.96
    रिडोर
    0.96
    ی
    0.95
    aue
    0.95
    Aslamualaikum
    0.95
    POSITIVE LOGITS
    ↵↵
    2.79
    ↵↵↵
    1.94
    ↵↵↵↵
    1.54
    ↵↵↵↵↵
    1.47
    ↵
    1.47
    <start_of_image>
    1.21
    ↵↵↵↵↵↵
    1.19
    ↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵
    1.13
    ↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵
    1.12
    ↵↵↵↵↵↵↵
    1.12
    Act Density 0.314%

    No Known Activations