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

    Neuronpedia

    INDEX
    Explanations

    base followed by identifiers

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    </td>
    -3.64
    .
    -2.83
     "
    -2.66
     封面
    -2.42
    ’
    -2.38
     didnt
    -2.38
     répand
    -2.28
    铩
    -2.19
    </h5>
    -2.11
    /"
    -2.09
    POSITIVE LOGITS
    Ⲉ
    2.61
     '-':
    2.41
    </strong>
    2.30
    ጁ
    2.25
    es
    2.22
    !”
    2.16
    同じく
    2.16
    𖥦
    2.14
    ଝ
    2.13
    蛧
    2.11
    Act Density 0.036%

    No Known Activations