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

    Neuronpedia

    INDEX
    Explanations

    cut, tear, break, remove

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    V
    0.85
    AR
    0.84
    ES
    0.83
    VE
    0.77
    ST
    0.76
    IC
    0.75
    W
    0.74
    Z
    0.73
    ۵
    0.73
    لی
    0.71
    POSITIVE LOGITS
     extracting
    0.68
     sót
    0.66
     ripping
    0.66
    at
    0.65
    اية
    0.64
     rád
    0.64
     tearing
    0.63
     cuttings
    0.62
     extractor
    0.62
     rip
    0.62
    Act Density 0.392%

    No Known Activations