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

    Neuronpedia

    INDEX
    Explanations

    couple relationship state

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     these
    -2.05
    /)
    -1.70
     those
    -1.64
     that
    -1.60
     this
    -1.57
     of
    -1.49
     we
    -1.49
    __)
    -1.48
    _));
    -1.46
    }'.
    -1.40
    POSITIVE LOGITS
    虽然
    1.50
    3
    1.42
    是怎么
    1.41
    雖然
    1.36
    5
    1.33
    ５
    1.33
    Then
    1.33
    至于
    1.30
    ПРИ
    1.29
     سپس
    1.29
    Act Density 0.024%

    No Known Activations