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

    Neuronpedia

    INDEX
    Explanations

    dividing into fragments or chunks

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    燗
    0.78
    ධ
    0.74
    贷
    0.73
    तूल
    0.69
    oleč
    0.67
    骋
    0.66
    芡
    0.66
    ర్శ
    0.66
     الوط
    0.65
     postureProxy
    0.65
    POSITIVE LOGITS
     fragments
    3.33
     fragment
    3.27
     fragmentation
    3.06
     fragmented
    3.03
     Fragment
    2.98
    fragment
    2.92
     chunks
    2.87
    Fragment
    2.84
     Fragments
    2.84
     chunk
    2.79
    Act Density 1.642%

    No Known Activations