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

    Neuronpedia

    INDEX
    Explanations

    contractions

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    1
    -1.52
     
    -1.40
     nggak
    -1.34
    ”
    -1.33
     &
    -1.30
     to
    -1.29
     затем
    -1.29
    ”.
    -1.27
     as
    -1.27
    </i>
    -1.23
    POSITIVE LOGITS
    Dont
    2.06
     wasnt
    1.84
     isnt
    1.70
    StoryboardSegue
    1.54
     Dont
    1.52
    itys
    1.51
    seurs
    1.48
    ugges
    1.48
     couldnt
    1.47
     میشود
    1.46
    Act Density 0.024%

    No Known Activations