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

    Neuronpedia

    INDEX
    Explanations

    :

    np_max-act · gemini-2.0-flash

    This neuron fires on metadata labels (e.g. “author,” “bibliography,” etc.) in academic‐paper headers.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     patri
    -0.07
     azi
    -0.07
    .Horizontal
    -0.07
     mountains
    -0.07
     harmless
    -0.07
     denn
    -0.06
    (pl
    -0.06
    esidir
    -0.06
    /md
    -0.06
    mary
    -0.06
    POSITIVE LOGITS
     acesso
    0.07
     Flip
    0.07
    ENSIONS
    0.07
     GDPR
    0.06
    PRIMARY
    0.06
    CSV
    0.06
    ble
    0.06
    έντρο
    0.06
     페이지
    0.06
     Signature
    0.06
    Act Density 0.004%

    No Known Activations