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

    Neuronpedia

    INDEX
    Explanations

    scaling resources up or down

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    校验
    0.45
    沉积
    0.45
     localhost
    0.44
     Conflict
    0.43
     Privacy
    0.42
     depos
    0.41
    矗
    0.40
    Privacy
    0.39
    선
    0.39
    Checksum
    0.39
    POSITIVE LOGITS
     scaling
    1.02
     Scaling
    0.92
    Scaling
    0.91
    scaling
    0.87
     scaled
    0.79
    Scaled
    0.74
    scaled
    0.69
     scal
    0.64
    scale
    0.63
     Scal
    0.63
    Act Density 0.010%

    No Known Activations