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

    Neuronpedia

    INDEX
    Explanations

    questions about X

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     with
    1.21
     to
    1.12
     and
    1.10
     a
    1.03
     in
    1.02
     along
    0.97
     from
    0.96
     that
    0.94
     on
    0.93
     into
    0.93
    POSITIVE LOGITS
     importance
    1.53
     biggest
    1.49
     effectiveness
    1.38
     latter
    1.37
     applicability
    1.36
     clearest
    1.35
     difference
    1.34
     plupart
    1.34
     impetus
    1.33
     efficacy
    1.32
    Act Density 0.133%

    No Known Activations