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

    Neuronpedia

    INDEX
    Explanations

    discover or notice

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    способ
    0.73
     possam
    0.73
     способны
    0.73
     способ
    0.70
    你可以
    0.69
    あなたの
    0.68
     possano
    0.68
     nonempty
    0.68
    capable
    0.68
     രാജ്യ
    0.67
    POSITIVE LOGITS
    发现
    1.04
    發現
    1.01
     discovered
    1.00
    พบ
    0.96
     noticed
    0.93
     발견
    0.92
     purchased
    0.92
     opted
    0.91
     bought
    0.90
     luck
    0.90
    Act Density 0.382%

    No Known Activations