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

    Neuronpedia

    INDEX
    Explanations

    describes states or degrees

    np_max-act-logits · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     Very
    0.54
     খুব
    0.53
     very
    0.52
    とても
    0.51
     muito
    0.50
    Very
    0.50
     очень
    0.48
     অত্যন্ত
    0.48
    very
    0.46
    很好
    0.46
    POSITIVE LOGITS
     likely
    0.73
     than
    0.67
     prevalent
    0.63
    likely
    0.63
     pronounced
    0.60
     akin
    0.59
     prone
    0.59
     attuned
    0.59
     frequent
    0.58
     stringent
    0.57
    Act Density 0.444%

    No Known Activations