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

    Neuronpedia

    INDEX
    Explanations

    fire hazard, resistance, firewall

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    _{\
    3.22
    _{*
    2.72
     gestation
    2.59
     nasty
    2.58
    ا
    2.54
     prized
    2.47
     pieno
    2.39
    isode
    2.35
    ̷
    2.34
     وش
    2.32
    POSITIVE LOGITS
    ために
    3.22
    🔥🔥
    3.21
    nze
    3.18
    ための
    3.03
     distinguishers
    2.82
    crackers
    2.68
    walls
    2.62
    𝘀
    2.49
    יות
    2.49
    ోతి
    2.47
    Act Density 0.061%

    No Known Activations