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

    Neuronpedia

    INDEX
    Explanations

    words ending in ance, ence, iance

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    sh
    -0.20
    set
    -0.19
    so
    -0.19
    ness
    -0.18
    shift
    -0.18
    ship
    -0.18
    son
    -0.18
    name
    -0.17
    rec
    -0.17
    reg
    -0.17
    POSITIVE LOGITS
    es
    0.21
    y
    0.18
    able
    0.18
    ously
    0.18
    ous
    0.17
    urs
    0.15
    ance
    0.15
    au
    0.14
    ois
    0.13
    e
    0.13
    Act Density 0.174%

    No Known Activations