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

    Neuronpedia

    INDEX
    Explanations

    tokens aren't words

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     counterpart
    0.39
     billion
    0.36
    स्टोन
    0.35
     விவர
    0.35
     dynam
    0.34
     experimental
    0.34
    exchange
    0.34
     incompar
    0.34
     exchange
    0.33
     Angry
    0.33
    POSITIVE LOGITS
    等多
    0.43
     fitting
    0.41
    不断的
    0.40
    Fitting
    0.40
     smug
    0.39
     adjectives
    0.38
     Pilgr
    0.38
     Phrases
    0.38
    ځ
    0.38
     phù
    0.38
    Act Density 0.004%

    No Known Activations