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

    Neuronpedia

    INDEX
    Explanations

    which after a comma

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    ว่า
    1.03
     rằng
    1.01
    ວ່າ
    0.95
     bahwa
    0.85
    ながら
    0.83
     أن
    0.82
     ότι
    0.82
     ว่า
    0.75
    ว่าจะ
    0.75
    போதும்
    0.74
    POSITIVE LOGITS
     which
    6.61
    which
    6.34
    Which
    5.71
     Which
    5.68
     WHICH
    5.24
     которой
    4.25
     которая
    4.19
     whiche
    4.19
     οποία
    4.19
     которые
    4.16
    Act Density 1.374%

    No Known Activations