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

    Neuronpedia

    INDEX
    Explanations

    a

    np_max-act · gemini-2.0-flash

    This neuron detects when the text is asking about or stating “relevance,” i.e. it flags relevance-evaluation language.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     áreas
    -0.07
    ега
    -0.07
    ınıza
    -0.06
     experiencia
    -0.06
     pools
    -0.06
    .dp
    -0.06
    _front
    -0.06
    aland
    -0.06
    CEEDED
    -0.06
    .IndexOf
    -0.06
    POSITIVE LOGITS
    (pipe
    0.07
    ृत
    0.07
    เศ
    0.06
     exagger
    0.06
     landlord
    0.06
    .ipv
    0.06
     जन
    0.06
    ]));
    0.06
    (',');↵
    0.06
     zpráva
    0.06
    Act Density 0.020%

    No Known Activations