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

    Neuronpedia

    INDEX
    Explanations

    Opinions/Feelings

    np_max-act · gemini-2.0-flash

    The neuron detects tokens that refer to people or social agents (words like "people", "who", "guys", "one", etc.).

    oai_token-act-pair · gpt-5-miniTriggered by @vetterc0
    New Auto-Interp
    Negative Logits
    CTYPE
    -0.06
    .NewReader
    -0.06
     notion
    -0.06
     gearbox
    -0.06
    _DIG
    -0.06
    emale
    -0.06
     yatırım
    -0.06
    酸
    -0.06
    рою
    -0.06
     bitcoins
    -0.06
    POSITIVE LOGITS
    .removeAttribute
    0.07
    RELATED
    0.07
     fwrite
    0.07
     Rein
    0.06
     overwritten
    0.06
    .oauth
    0.06
     assigned
    0.06
    references
    0.06
    DATED
    0.06
     altitude
    0.06
    Act Density 0.748%

    No Known Activations