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

    Neuronpedia

    INDEX
    Explanations

    extracts and extraction processes

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     returnPacket
    0.69
     özellikle
    0.68
    emptyDict
    0.62
    <0x94>
    0.62
     информацию
    0.59
     construit
    0.58
    -'+
    0.58
    0.58
     второй
    0.58
     เข
    0.57
    POSITIVE LOGITS
    wood
    0.63
     extracts
    0.63
    ties
    0.60
    degree
    0.59
     extract
    0.59
     Extract
    0.58
    eyes
    0.56
    som
    0.55
    ase
    0.54
    aus
    0.54
    Act Density 0.009%

    No Known Activations