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

    Neuronpedia

    INDEX
    Explanations

    SQLAlchemy `json_agg`

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    ançais
    0.35
    ativement
    0.33
    /*
    0.33
     por
    0.32
     due
    0.32
     profondément
    0.32
     sprinting
    0.32
     jeweils
    0.31
    σωπ
    0.31
     grants
    0.31
    POSITIVE LOGITS
     то
    0.50
     коне
    0.47
    ône
    0.46
    agons
    0.45
     mús
    0.44
     название
    0.44
    tube
    0.42
    ᱓
    0.42
     तो
    0.42
     মুক্তিব
    0.42
    Act Density 0.192%

    No Known Activations