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

    Neuronpedia

    INDEX
    Explanations

    Did, Interestingly, But

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     خواهد
    0.76
     devra
    0.74
     erfolgt
    0.69
    后的
    0.69
    での
    0.66
     erfolgen
    0.66
    不算
    0.63
    への
    0.63
     محک
    0.62
     соответствует
    0.62
    POSITIVE LOGITS
    Did
    2.58
     Did
    2.46
     did
    2.40
    did
    2.15
     DID
    1.76
     didn
    1.74
     Interestingly
    1.73
    Interestingly
    1.66
    你知道
    1.65
    DID
    1.59
    Act Density 0.669%

    No Known Activations