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

    Neuronpedia

    INDEX
    Explanations

    deductions and SALT limitations

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     abierta
    0.39
     پی
    0.38
     Printers
    0.38
    órias
    0.37
     bisnis
    0.37
     เดือน
    0.37
     frut
    0.36
     Hala
    0.36
     Hap
    0.36
     frutas
    0.35
    POSITIVE LOGITS
     ded
    0.47
    ded
    0.47
     SALT
    0.46
     limitations
    0.45
    denly
    0.45
    salt
    0.43
     Cuomo
    0.42
     deduct
    0.42
    itons
    0.41
     salt
    0.41
    Act Density 0.008%

    No Known Activations