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

    Neuronpedia

    INDEX
    Explanations

    describing skill proficiency

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    nél
    1.94
    ts
    1.70
    <0xB2>
    1.64
    il
    1.63
    nz
    1.60
    ewnętr
    1.58
    ለያዩ
    1.54
    <0xA5>
    1.52
    었던
    1.49
    et
    1.48
    POSITIVE LOGITS
    Το
    1.64
    ২
    1.64
    ﻣ
    1.55
    Με
    1.48
     sheer
    1.47
    ﻧ
    1.45
    Πα
    1.44
    Го
    1.42
    要知道
    1.42
    ਿ
    1.42
    Act Density 0.005%

    No Known Activations