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

    Neuronpedia

    INDEX
    Explanations

    key differences and comparisons

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    бль
    0.44
    ほと
    0.43
     malignant
    0.39
     spectrom
    0.39
    ুয়ার
    0.39
    armes
    0.38
     noirâtre
    0.38
     flancs
    0.37
     arco
    0.37
     piety
    0.37
    POSITIVE LOGITS
    ---|
    0.43
    iam
    0.39
    তাকে
    0.38
    ⋮
    0.37
     Rate
    0.37
     ;
    0.36
     yapılan
    0.36
    aram
    0.36
     IPA
    0.36
     BBB
    0.36
    Act Density 0.009%

    No Known Activations