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

    Neuronpedia

    INDEX
    Explanations

    rather than or instead of

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     although
    0.43
     नक्की
    0.41
     though
    0.38
    Although
    0.38
     choć
    0.36
     אך
    0.36
     ranging
    0.34
     Apache
    0.34
     allerdings
    0.34
    ᕝ
    0.34
    POSITIVE LOGITS
    而不是
    0.83
     rather
    0.78
    rather
    0.76
    而非
    0.66
     instead
    0.64
    instead
    0.64
     बजाय
    0.63
     chứ
    0.60
     piuttosto
    0.58
     plutôt
    0.57
    Act Density 0.166%

    No Known Activations