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

    Neuronpedia

    INDEX
    Explanations

    go beyond or go wrong

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     goes
    0.42
     गई
    0.41
     sete
    0.40
     went
    0.40
    went
    0.40
     गए
    0.39
     Goes
    0.38
    keley
    0.37
     जातो
    0.37
    goes
    0.37
    POSITIVE LOGITS
     wrong
    0.47
     overboard
    0.43
     mad
    0.41
     SPL
    0.40
     unnoticed
    0.37
    தமாக
    0.37
    VERN
    0.37
    حاب
    0.37
    VerFile
    0.37
     Shil
    0.36
    Act Density 0.033%

    No Known Activations