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

    Neuronpedia

    INDEX
    Explanations

    variable assignment

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     experimentally
    1.78
    ${
    1.73
     actu
    1.71
    ве
    1.69
    ɪ
    1.65
    écie
    1.63
     waxaa
    1.60
     hust
    1.60
     Canning
    1.59
     Besonders
    1.49
    POSITIVE LOGITS
    ל
    1.72
    ுகிறார்கள்
    1.71
    তে
    1.70
     `=
    1.67
    𓏸
    1.64
    withstanding
    1.63
    𝐭
    1.58
    ]`
    1.55
    𝐂
    1.55
    तीर्ण
    1.53
    Act Density 2.425%

    No Known Activations