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

    Neuronpedia

    INDEX
    Explanations

    code placeholders

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     feels
    0.98
     armies
    0.87
     garlic
    0.84
     bones
    0.84
     army
    0.84
     feelings
    0.83
     pumpkins
    0.83
     Army
    0.82
     navy
    0.82
     soldiers
    0.82
    POSITIVE LOGITS
     пове
    0.64
     постав
    0.64
    webhook
    0.63
    ông
    0.62
     вики
    0.61
     নিয়
    0.61
    ミア
    0.60
     ஆப
    0.60
    ोल
    0.60
    या
    0.59
    Act Density 0.008%

    No Known Activations