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

    Neuronpedia

    INDEX
    Explanations

    xml tags and math notation

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     devolución
    0.53
    掎
    0.53
     đào
    0.52
     loja
    0.52
     руба
    0.52
     mân
    0.51
     возле
    0.51
    埝
    0.51
    蒉
    0.51
    ắn
    0.50
    POSITIVE LOGITS
    ↵
    0.60
     \
    0.56
    0.56
     whose
    0.55
    L
    0.55
     <
    0.54
     trivially
    0.54
     {
    0.53
                    
    0.53
     functor
    0.52
    Act Density 0.009%

    No Known Activations