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

    Neuronpedia

    INDEX
    Explanations

    it followed by 'hard' or 'would'

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     fleste
    1.40
     fornisce
    1.02
     méridionale
    1.02
     виправивши
    1.01
     provide
    0.97
     bisogno
    0.97
    ート
    0.95
    instances
    0.95
    SAND
    0.94
     Portanto
    0.93
    POSITIVE LOGITS
    ли
    1.32
    a
    1.28
    ことで
    1.25
    p
    1.25
    说
    1.19
    И
    1.13
    こと
    1.13
    О
    1.12
    ﺐ
    1.12
    م
    1.11
    Act Density 0.333%

    No Known Activations