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

    Neuronpedia

    INDEX
    Explanations

    format and technology

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     während
    0.46
     dur
    0.44
     durante
    0.42
     DUR
    0.41
     suspensions
    0.41
     během
    0.41
     supprim
    0.40
     raíz
    0.40
     geh
    0.40
    ibilité
    0.39
    POSITIVE LOGITS
    ABCD
    0.45
    Paying
    0.44
     Paying
    0.44
    Analyzing
    0.43
    Combine
    0.42
    objective
    0.42
    Against
    0.40
    Keep
    0.40
    Architect
    0.40
    combine
    0.40
    Act Density 0.001%

    No Known Activations