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

    Neuronpedia

    INDEX
    Explanations

    Residual computation/coding

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Negative Logits
    سر
    -0.09
    enny
    -0.08
     superst
    -0.08
    _ai
    -0.07
    ச்ச
    -0.07
    ثمر
    -0.07
     transc
    -0.07
     safer
    -0.07
     safe
    -0.07
     العم
    -0.07
    POSITIVE LOGITS
     residual
    0.11
    Residual
    0.10
    残
    0.09
     unmet
    0.09
     Resid
    0.09
    /problem
    0.08
     discrepancy
    0.08
     untreated
    0.08
     discrepancies
    0.08
     fé
    0.08
    Act Density 0.002%

    No Known Activations