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

    Neuronpedia

    INDEX
    Explanations

    First-person perspective

    np_max-act · gemini-2.0-flash

    first-person self-referential statements that express intent, desire, or ability, especially when paired with modal/auxiliary verbs in user queries.

    oai_token-act-pair · gpt-5Triggered by @vetterc0
    New Auto-Interp
    Negative Logits
     besch
    -0.08
    ЛЕ
    -0.07
    [hash
    -0.07
     לראש
    -0.07
    ผ
    -0.07
    .land
    -0.07
    ','#
    -0.07
    .),
    -0.07
    饮用
    -0.07
    여
    -0.07
    POSITIVE LOGITS
    煴
    0.08
     outcome
    0.08
    ço
    0.07
    oklyn
    0.07
    iktig
    0.07
    “When
    0.07
    😬
    0.07
     analyzes
    0.07
    BD
    0.07
    FD
    0.07
    Act Density 0.103%

    No Known Activations